[INST] - Verify Multiple CallHome config on Connector [/INST] : ESMBD 32070 - TAC - Connectors Stuck in delete In progress State > 8hours in Director. Prechecks - How to identify , if callhome button is used on multiple connectors in director logs. grep -i “ctr?actn=clhome” app1_* <<< lists connector on which callhome option was used. YOU MUST RUN ON EVERY/ALL CONNECTORS IN MANO SYSTEM APPLICABLE ONLY IN 19.1.1 SYSTEM & PRIOR. PLEASE REMOVE CALLHOMESCRIPTS & BUNDLE FROM MANO ONCE [INST] IS EXECUTED. Option1 - Remove multiple entries of callhome config from the connector 1.Login into controller. 2.Download CleancallhomeScripts.zip to /tmp 3.Unzip the bundle 4.Execute chmod 777 on get_callhome & clean_duplicate_callhome.sh scripts. 5.Update the controller ip in get_callhome & clean_duplicate_callhome.sh scripts controller_ip="" 6.Execute ./get_callhome.sh | tee /tmp/callhomeconfig.txt 7.Output will list connectors with mutltiple callhomeconfig will be listed as below Callhome Count of connector with UID :: " " :: is 2, One of the config entry will listed as \"name\": \"CA\" 8.To remove Multiple callhome Entries , Execute ./clean_duplicate_callhome.sh | tee /tmp/cleancallhomeconfig.txt 9. After step 8 is completed , check netstat -anp | grep 4334 lists connectors in callhome esblaished state 10. If step 9 doesnt list callhome in established state. Login to connector with security credentials & apply ovnchallenge,ovn response to root shell prompt Restart processes in connector by running below command in the connector shell (not in the CLI prompt)- #systemctl restart SysRepoSync sysrepo netopeer 11.Verify call home Session is established in Controller through netstat -anp | grep 4334. With a gap of 10 secs, to double confirm on consistent State. [INST] - To correct callhome/oper down status in Director when connector is reachable [/INST] : ESMBD 32070 -how to identify callhome Disconnected State in dir is due "unable to connect Issue in controller" This issue may be observed only during controller reboot. how to identify : a. In director , callhome status shows Oper down/disconnected b. run "netstat -anp | grep 4334 | grep " for 3 times with a gap of 10 seconds on controller. observe the connection status remained "established" and the remote end-point port should not be changing in all the runs. netstat o/p must have exact number as connectors in Director . 10 connectors , 10 netstat unique callhome sessions in established state c. Ping connector mgmt IP from controller to double confirm reachability .If reachable apply below workaround d. Check Step 6 of ESMBD-32220 if any multiple callhome config for all connectors. Reason- Multiple callhome config in different connectors can also impact. If all above conditions are met & confirms "unable to connect Issue in controller" workaround: Initiate "Reboot" on that connector from Director e. If above workaround is impacting datatraffic Try by deleting & adding callhome config on connector available in :CallHome_Scripts.7z The steps are : * Login into controller. * Download callhomeScripts.zip [getcallhome.sh,addcallhome.sh,deletecallhome.sh] to /tmp * Unzip the bundle * Execute chmod 777 on all scripts [getcallhome.sh,addcallhome.sh,deletecallhome.sh] * Get the list of callhome config from a specific connector , use get callhome.sh script & Update Security Credentials in the script * Verify if the list contains callhome config with entry as \"name\": \"CA\" << This indicates CallHome button Signature from Director UI * To delete this entry, remove deletecallhome.sh update security credentials in deletecallhome.sh * Re-run getcallhome.sh ,to check all callhome config is removed * Add callhome config >Update UID echo "curl -k -X PUT https://$CONNECTOR_IP:5000/vse/api/v1.0/config/working/callhome/-H 'cache-control: no-cache' -H 'content-type: application/json' -H \"x-auth-token: $AUTHTOKEN\" -d '{\"max-wait\": \"30\", \"name\": \"\",\"idle-timeout\": \"300\",\"reconnect-maxattempts\": \"4\",\"reconnect-timeout\": \"1\",\"connection-type\": \"persistent\",\"endpoints\": [{\"port\": \"4334\",\"endpoint\": \"ed2\",\"name\": \"\",\"address\": \"$CONTROLLER_IP\"}],\"max-attempts\": \"3\",\"transport\": \"ssh\",\"reconnect-strategy\": \"last-connected\"}' " > /tmp/add_call_home_$CONTROLLER_IP >Update security credentials LOGIN_OUTPUT=$(curl -k -X PUT https://$CONNECTOR_IP:5000/vse/api/v1.0/login/ -H 'cache-control: no-cache' -H 'content-type: application/json' -d '{"username" : "security","password" : "security"}') * Run addcallhome Script -u connectorip -c * Verify whether there is only one entry of callhome config by running the get_callhome.sh script * Login to connector with security credentials & apply ovnchallenge,ovn response to root shell prompt * Restart processes in connector by running below command in the connector shell (not in the CLI prompt)-#systemctl restart SysRepoSync sysrepo netopeer * Verify Single call home Session is established in Controller through netstat -anp | grep 4334. With a gap of 10 secs, to double confirm on consistent State Verify if State changes in Director UI [INST] - To remove Connector Delete In Progress Stuck Entries in Controller [/INST] : ESMBD 32070 - TAC - Connectors Stuck in delete In progress State > 8hours in Director. 1. YOU MUST ENSURE NO MULTIPLE CALLHOME CONFIG ON CONNECTORS as per Question ESMBD-32220 2. REQUIRE TO USE ONLY WHEN CONNECTOR ENTRIES ARE IN STUCK STATE in DIRECTOR,CONTROLLER. To check the stuck entries in the controller database, try the following from the controller terminal – (substitute the controller IP appropriately) curl -X GET 'https://:8443/restconf/operational/adva-ensemble-controller:ensemble-controller/connector-config:connector-configs-deletion-status' -H 'Authorization: Basic YWRtaW46YWRtaW4=' -H 'Accept: application/json' -k O/P is UID list with <> Above o/p indicates , deletion is stuck due to unavailable threads in Controller 3. Collect showTech -m true 4. Reboot the controller form terminal ,unblocks threads hung state. sudo reboot 5. Use the following to cleanup the stuck entries (forcefully cleanup) curl -k -X POST 'https://:8443/restconf/operations/baseservice:delete-stale-entries'-H 'Authorization: Basic YWRtaW46YWRtaW4=' -H 'Content-Type: application/json' -H 'Accept: application/json' -d '{ "input": { "stale-entries": [ "/adva-ensemble-controller:ensemble-controller/connector-config:connector-configs-deletion-status" ], "data-store": "operational" }}' 6. From the director terminal , run deleteConnector.sh -c -u (path: /opt/adva/nc/aos-smartwan-installer/bin) o/p - 200 Ok ,also monitor Controller karaf logs if preconfigs are successfully cleaned up & in Orchestrator ,cloud/ Physical Device /Service is cleaned up. [INST] - All connectors in director reporting call home oper down at same time. [/INST] : PATCH INSTALL CONFIRMATION: Once Patch for the callhome issue is applied on Customer prod Controller (single or cluster). Verify by going to the directory: cd /opt/controller/system/com/adva/ensemble/callhome-impl/20.2.1.4/ do an "ll" to see the files and timestamps. You should see an .jar file with the timestamp of the day and time the patch was installed. MOP to solve the issue of oper status of the connectors shown down in Director- Issues reported by Lumen, Singtel & Granite Ref https://polarion.advaoptical.com/polarion/#/project/ESMBD/workitem?id=ESMBD-32070 https://polarion.advaoptical.com/polarion/#/project/ESMBD/workitem?id=ESMBD-35850 conditions - 1. MOP Applicable only on 19.1.1 MANO Setups 2. Procedure applicable if 19.1.1 Directors Patch P1 or Controller Patch P1 is not applied. Impact - 1. POST ZTP on Director , Oper Status of ALL connectors are reported as down in director UI 2. POST ZTP on Controller, Call home Connections are in SYNC_REC or CLOSE_WAIT for all connectors. This can be noticed after controller reboot. - Verify through netstat -anp | grep 4334 SIGNATURES TO CONFIRM ON CONTROLLER DURING LIVE SESSION: 1. netstat -anp | grep 4334 - o/p must display many connectors in Close_wait state 2. grep -i "Keepalive RPC failed with error: [RpcError [message=Session closed, severity=ERROR, errorType=TRANSPORT, tag=operation-failed, applicationTag=null, info=null, cause=null]]" karaf.log 3. on Director - All connectors must be reported as callhome oper down. 4. Verify if threads are in hung state. => This step must be done while the issue exists. Logs won't provide the real status of these threads, it must be done during the live troubleshooting session. [ensemble@controller scripts]$ jps 1738 Main -----> controller pid 14781 Jps Redirect thread dump to a file ensemble@controller scripts]$jstack 1738 > thread.txt open the file through a vi editor & find all - "NioProcessor-" string Check no of occurences of "NioProcessor-X" [X -1, 2,3,4,5] threads & state of the thread. if 3/5 are in Blocked/waiting state for longer period . Then conclusion threads are in Hung state.samples : "NioProcessor-1" #455 prio=5 os_prio=0 tid=0x00007f7b7801b800 nid=0xbc8 runnable [0x00007f7b2e25d000] java.lang.Thread.State: RUNNABLE <<<<< BLOCKING ===> important to note the "tid" must be the same across the 5 threads. 5. Repeat step 4 for at least 5mins with the gap 3 secs & redirect to diff file name each time. If 3/5 are in Blocked state for longer period . Then conclusion threads are in Hung state 6. Take a controller show tech -d true VERIFICATIONS TO APPLY MOP: 1. What to monitor ? Run below command on controller sudo lsof -u ensemble | wc -l (takes 120 sec or more or count >400000 or if all Connectors are shown operationally down on Dir – execute the MOP below) 2. Check call home connections on controller whether in Established State or SYNC_REC/CLOSE_WAIT netstat -anp | grep 4334 3. Suppose Director P1,P2 patches are not applied & callhome button is executed on connectors from Director UI Verify in director logs cd /var/opt/adva/log/aos/nc/ grep -i “ctr?actn=clhome” if o/p yields Connector Entries e.g 2022-01-18T13:38:30.251Z|RESTAPIHandler|INFO|1fb3c130a9dc5983||Got POST call on /mit/md/61549adfefe637683295eaf2/me/61a780c8efe63768329bbf8a/ctr?actn=clhome 4. Issue can be observed on Scale Setup or 19.1.1 MANO system soaked for longer time 5. Director UI should report Oper down on all connectors post ZTP. Procedure is as follows - STEP1: Disable the callhome-recovery-feature(this is only one-time). Block callhome connections upon controller retsart. Once controller is started successfully,unblock the callhome Port sudo lsof -u ensemble | wc -l (takes 120 sec or more or count >400000 or if all Connectors are shown operationally down on Dir – execute below) Login to the controller using ‘ensemble’ user sudo systemctl stop controller //wait for controller to stop, check using "sudo systemctl status controller", nodeStatus sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 -j DROP Update the following attributes in $CONTROLLER_HOME: /opt/controller/etc/com.adva.ensemble.controller.cfg, with the values given below - callhome-recovery-wait-time=0 waiting-time-to-receive-callhome=30 Save the file sudo systemctl start controller //wait for controller to start check using "sudo systemctl status controller", nodeStatus //should show all services as up //wait for 2 min – this is necessary Make sure all nodes are up: grep -ir "aye captain" /opt/controller/data/log/karaf.log sudo iptables -D INPUT -i eth0 -p tcp --destination-port 4334 -j DROP sudo iptables -D INPUT -i eth0 -p tcp --destination-port 4334 --source-port 40000:60000 -j DROP If there are any rules other than the expected, please use the command "sudo iptables -L --line-numbers -n" to clean up and remove unwanted rules. Reverify if callhome sessions are establishing & on director UI,Oper up shown on all connectors in director STEP 2: This step is ** Applicable only & only if callhome button is exercised else not required. Verify the traces & proceed. One of the two options below can be used to remove the multiple callhome entries in connector- Verify in director logs cd /var/opt/adva/log/aos/nc/ grep -i “ctr?actn=clhome” if o/p yields Connector Entries Option1 - Remove multiple entries of callhome config from the connector 1.Login into controller. 2.Download CleancallhomeScripts.zip to /tmp 3.Unzip the bundle 4.Execute chmod 777 on get_callhome & clean_duplicate_callhome.sh scripts. 5.Update the controller ip in get_callhome & clean_duplicate_callhome.sh scripts controller_ip="" 6.Execute ./get_callhome.sh | tee /tmp/callhomeconfig.txt 7.Output will list connectors with mutltiple callhomeconfig will be listed as below Callhome Count of connector with UID :: " " :: is 2, One of the config entry will listed as \"name\": \"CA\" 8.To remove Multiple callhome Entries , Execute ./clean_duplicate_callhome.sh | tee /tmp/cleancallhomeconfig.txt 9. After step 8 is completed , check netstat -anp | grep 4334 lists connectors in callhome esblaished state 10. If step 9 doesnt list callhome in established state. Login to connector with security credentials & apply ovnchallenge,ovn response to root shell prompt Restart processes in connector by running below command in the connector shell (not in the CLI prompt)- #systemctl restart SysRepoSync sysrepo netopeer 11.Verify call home Session is established in Controller through netstat -anp | grep 4334. With a gap of 10 secs, to double confirm on consistent State. Option2 – Start a fresh ZTP on the connector. Also need to make sure callhome option is not used via Director. : #1 by Sourabh Pandey on 2021-02-24 09:37 Vikram, Krishna, [INST]: Allows to free up disk space on MANO Orchestrator component [/INST] : Tracks MOPdeliverable, procedural steps and updated script. This MOP contains a script that will free up the disk space used by the File Descriptors which are deleted but still open. Script is attached Comments : #1 by Tatiana Dietze on 2020-08-05 01:15 Please provide detailed instructions to run provided script. #2 by Sivasankar Ramalingam on 2020-08-06 14:57 Problem statement: The used disk space detail in the Orchestrator VM does not match with the actual space occupied by the Files and Folders in the filesystem. Possible root cause: This could be due to an issue in the Orchestrator log rotation mechanism which leaves the File Descriptors(FD) in the open state though they are deleted. How to identify the problem: Run "sudo lsof +L1" command to check if there are any open file descriptors still exist which are deleted. If it returns an output as shown in the below example with a list of files deleted but still in the open state. Those deleted files need to be cleared using the clear_deleted_fds.sh script. [ensemble@siva-standalone-orch-two ~]$ sudo lsof +L1 COMMAND PID USER FD TYPE DEVICE SIZE/OFF NLINK NODE NAME mysqld 6989 mysql 7u REG 253,0 0 0 264696 /tmp/ib7A3Vpy (deleted) mysqld 6989 mysql 8u REG 253,0 0 0 264729 /tmp/ibo3c2KN (deleted) mysqld 6989 mysql 9u REG 253,0 187 0 264730 /tmp/ibLpx852 (deleted) mysqld 6989 mysql 10u REG 253,0 0 0 264731 /tmp/ibfMR4Rx (deleted) mysqld 6989 mysql 14u REG 253,0 1196 0 264732 /tmp/ibqekIkN (deleted) java 27894 ensemble 1w REG 253,2 3173657158 0 262208 /var/log/ensemble/eso-core/orchestrator.log.1 (deleted) java 27894 ensemble 2w REG 253,2 3173657158 0 262208 /var/log/ensemble/eso-core/orchestrator.log.1 (deleted) java 27965 tomcat 104w REG 253,0 201475 0 786433 /usr/share/tomcat/osa.log.1 (deleted) [ensemble@siva-standalone-orch-two ~]$ The orchestrator.log.1 is ~3GB in size which is deleted but still listed in the open File Descriptor list and holds 3 GB of disk space. How to rectify the issue: 1. Run the "df -h" command to check the current disk utilization detail. 2. Copy the clear_deleted_fds.sh script to /opt/ensemble/eso-core/bin/scripts/ directory 3. Change the file permission and ownership by running the below commands. sudo chown ensemble:ensemble /opt/ensemble/eso-core/bin/scripts/clear_deleted_fds.sh sudo chmod +x /opt/ensemble/eso-core/bin/scripts/clear_deleted_fds.sh 4. Execute the script as shown below. /opt/ensemble/eso-core/bin/scripts/clear_deleted_fds.sh 5. Run "df -h" command to check the disk utilization detail after the clearing the deleted files. Note: Though the disk space is recovered, the 'lsof +L1' command may still return the same output until the eso-core service is restarted. #3 by Tatiana Dietze on 2020-08-11 16:41 This script helps to clean up deleted entries from the /var logs. Important: It won't make space for the /opt/ensemble/eso-core/vnf_images directory. #4 by Tatiana Dietze on 2020-08-30 11:26 This is not for the /opt/ensemble/eso-core/vnf_images directory. Another procedure is in progress to address space for the vnf images. [INST]: Service Creation Failure - Stuck in "Creating Virtual Networks" [/INST] : Tracks MOP deliverable, procedural steps and updated script. MANUAL COMMANDS TO FIX THE LOCK ISSUE: 1) Login to the DB : [ensemble@orchestrator ~]$ sudo mysql -u root Welcome to the MariaDB monitor. Commands end with ; or \g. Your MariaDB connection id is 524052 Server version: 10.1.19-MariaDB MariaDB Server Copyright (c) 2000, 2016, Oracle, MariaDB Corporation Ab and others. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. 2) Select osa database MariaDB [(none)]> use osa; Reading table information for completion of table and column names You can turn off this feature to get a quicker startup with -A Database changed MariaDB [osa]> select * from distributed_lock; +----+-----------------+------------------------+---------------------+----------+---------------------+---------------------+---------+ | id | eso_instance_id | eso_instance_thread_id | lock_name | owner_id | create_date | update_date | version | +----+-----------------+------------------------+---------------------+----------+---------------------+---------------------+---------+ | 1 | 3 | 49934003 | getCIDRPool | NULL | 2020-05-22 18:19:38 | 2020-07-22 21:58:20 | 871 | | 2 | NULL | NULL | associateSSHKey | NULL | 2020-05-22 18:19:38 | 2020-07-27 22:15:54 | 64 | | 3 | NULL | NULL | remediationPhase | NULL | 2020-05-22 18:19:38 | 2020-06-03 18:56:39 | 4 | | 4 | NULL | NULL | initializationPhase | NULL | 2020-05-22 18:19:38 | 2020-05-22 18:19:38 | 0 | +----+-----------------+------------------------+---------------------+----------+---------------------+---------------------+---------+ 4 rows in set (0.00 sec) 3) Update the distributed_lock MariaDB [osa]> update distributed_lock set eso_instance_id=NULL, eso_instance_thread_id=NULL; 4) Verification of lock release :Please check eso_instance_id, eso_instance_thread_id values to be NULL for the lock_name "getCIDRPool" in the distributed_lock table MariaDB [osa]> select * from distributed_lock; +----+-----------------+------------------------+---------------------+----------+---------------------+---------------------+---------+ | id | eso_instance_id | eso_instance_thread_id | lock_name | owner_id | create_date | update_date | version | +----+-----------------+------------------------+---------------------+----------+---------------------+---------------------+---------+ | 1 | NULL | NULL | getCIDRPool | NULL | 2020-05-22 18:19:38 | 2020-07-22 22:58:20 | 871 | | 2 | NULL | NULL | associateSSHKey | NULL | 2020-05-22 18:19:38 | 2020-07-27 22:15:54 | 64 | | 3 | NULL | NULL | remediationPhase | NULL | 2020-05-22 18:19:38 | 2020-06-03 18:56:39 | 4 | | 4 | NULL | NULL | initializationPhase | NULL | 2020-05-22 18:19:38 | 2020-05-22 18:19:38 | 0 | +----+-----------------+------------------------+---------------------+----------+---------------------+---------------------+---------+ 4 rows in set (0.00 sec) 5) Exit the DB CLI by pressing CTRL+Dor RUN THE SCRIPT (FILE ATTACHED) AND INSTRUCTIONS OR COMMENT #6. troubleshooting report : 1. login as admin to vrouter or sec-gw VEC_TAC_MANO_20_1_1_SEC_GW> cmd troubleshooting-report new Gathering information. This may take some time... Saved into /var/lib/yams/troubleshooting-reports/2022-03-24_14-01-06.tgz VEC_TAC_MANO_20_1_1_SEC_GW> Transfer /var/lib/yams/troubleshooting-reports/2022-03-24_14-01-06.tgz using scp * Using SCP : from root login in vrouter to pppoe server tmp location * Using SCP : from root login in sec-gw BGP * show bgp ipv4 unicast vrf mano * show ipv4-routes vrf mano * show bgp ipv4 unicast summary vrf mano * show state / vrf mano routing bgp Reachability/Traffic * show interface throughput vrf mano * cmd ping vrf mano 60.60.60.2 source 10.100.100.100 * cmd show-traffic ens160 * cmd ping vrf main or mano destip License * cmd license refresh * show license IKE * show ike ike-sa details vrf main Log * show log service ike vrf main * show log service license Performance Statistics * KPI metrics - https://doc.6wind.com/turbo-router-3/latest/turbo-router/user-guide/cli/monitoring/kpi.html?highlight=grafana Check if there are any IKE session drops, TCP States * show state system network-stack conntrack * ip netns exec mano ethtool -S ens5 * tcpdump -i any "tcp[tcpflags] & (tcp-syn|tcp-fin) != 0" [INST] - Director memory usage above 80% [/INST] : Polarian ESMBD-36234 How to monitor Director Memory usage 1. Throu monitor_dir.sh shared in MOP ESMBD-31569 MOP to apply a cap on the database cache – 1. Login to director VM 2. aos_stop 3. cd /opt/adva/nc/mongodb/conf/ 3.a sudo cp mongodb.conf backup_mongodb.conf 4. sudo vim mongodb.conf 5. modify this below block from storage: dbPath: /opt/adva/nc/mongodb/data/ to storage: dbPath: /opt/adva/nc/mongodb/data/ wiredTiger: engineConfig: cacheSizeGB: 3 6. save the file 7. aos_start [INST] - Collect Network Metrics(MTR,Traceroute,Ping,TCP States, Bandwidth),tcpdump on any any given system [/INST]: Polarian ESMBD-35850 monitor_NetworkMetrics.sh script helps to collect preliminary network data , which can be used to determine Packet loss, latency to Destined IPs from source ips. * Traceroute * Mtr * Ping64 bytes Captures help to an extent of evaluating MANO hosted Environment and can be applied on any system if monitoring is required 1. Install mtr package on source machine if pkg is unavailable.sudo yum install mtr2. Download monitor_Networkmetrics zip script,ips.txt files to source machines untar the zip file Provide permission to script & file Example - Controller- monitor_Networkmetrics..7z to /home/ensemble path 6wind - monitor_Networkmetrics.7zips.txt to / path chmod 777 monitor_Networkmetrics.sh ips.xt 3. Edit ips.txt file and mention the list of destination Ips to which metrics needs to collected from source machine.4. Schedule execution through cronjob on hourly/once in day (or as per need basis) or can be used for ONE time data collection as well.5. Tcpdump in addition sudo tcpdump -i any -w systemname.pcap "tcp[tcpflags] & (tcp-syn|tcp-fin) != 0" -nn -v6. iperf capture also helps to analyse bandwidth7. Remote side(connectors) -> Controller same test need to be done.8. ss - socket statistics similar to netstat Usage - * ./monitor_NetworkMetrics.sh * sudo tcpdump -i any -w .pcap "tcp[tcpflags] & (tcp-syn|tcp-fin) != 0" -nn -v >>> Captures tcp traffic . Can be fine tuned as per need basis. * Iperf Usage Example- //Enable Iperf on connector firewall-profile default firewall-port 5201 enable all incoming back iperf vrf "default" -s -p 5201 Run this from source machine - iperf3 -c 10.1.18.178 -p 5201 4.Connector - sudo tcpdump -i vti-ipsec1 -w connip.pcap "tcp[tcpflags] & (tcp-syn|tcp-fin) != 0" -nn -v 5.ss -s -à summary 6.ss -A tcp --- ALL TCP states Results - * o/p Data from all captures can be analysed to Evaluate Packet Loss, Network Latency , Bandwidth, TCP States(Stale, incomplete Handshake) * XX_mtr.csv * XX_traceroute.csv * XX_ping64.csv * Tcpdump pcap files on viceversa * Iperf data * ss statistics o/p from the source machine [INST] : Verify the No of Event Records in Director DB [/INST] : Verifying this will be applicable only after 24 hours from setting log purging over director * Run the cluster status ,identify Primary DB node * Run below query from Primary DB node * sudo /opt/adva/nc/mongodb/bin/mongo admin -u root -p 4netMGMT * use nc * db.log_comlog.stats() From the above o/p ,verify count in below section { "ns" : "nc.log_comlog", "size" : 7938526464, "count" : 3852311, >>>>>>this indicates 3.8 M Event records in comlog table "avgObjSize" : 2060, "storageSize" : 5100269568, "capped" : false, "wiredTiger" : { ……} [INST] : DI(5G_PPPoE_GPON ) - IKE going down post ZTP Troubleshooting [/INST] : Network Topology Debug Summary 1. Check if next hop is reachable 2. Check if WAN ip is reachable/assigned 3. Check Routes on vrouter 4. Check outstanding faults 5. Check if manual changes are done on vrouter to match 5G 6. Check the exact sequence done? 7. Is it 5G from day0 or switch b/w 5G & pppoe 8. Collect Pcaps 9. ZTP activations time 10. Check the NAT rules on vrouter 11. If IP is blacklisted in Etisalat network? 12. Check interfaces all Oper up 13. Check if physical connection is on right port in customer premises [INST] : Mano 2.0 - ShowTech files triage [/INST]: MANO 2.0 showTech - by default will have files from all components Download the showTech Untar the showTech file 1. To check current version of all pods running with : showtech_2024.02.20_08.56.45.tar\showtech_2024.02.20_08.56.45\install_logs Check the version file which captures pod version running with 2. To check the director logs : Most of the dir log triage remains same as MANO 1.0 Only the log path has changed showtech_2024.02.20_08.56.45.tar\showtech_2024.02.20_08.56.45\director Any exceptions wrt Operations can be refered in below logs This folders contains : app_c1, app_scala 3. Log path showtech_2024.02.20_08.56.45.tar\showtech_2024.02.20_08.56.45\install_logs Cloud property file : showtech_2024.02.20_08.56.45.tar\showtech_2024.02.20_08.56.45\install_logs\conf MANO Installation : showtech_2024.02.20_08.56.45.tar\showtech_2024.02.20_08.56.45\install_logs\log & nohup ShowTech log : showtech_2024.02.20_08.56.45.tar\showtech_2024.02.20_08.56.45\install_logs\showtech MANO Version : showtech_2024.02.20_08.56.45.tar\showtech_2024.02.20_08.56.45\install_logs\version 4. Pod Logs : showtech_2024.02.20_08.56.45.tar\showtech_2024.02.20_08.56.45\pods Egs : Controler K8 pods , Kube-system , Polling agent ,polling master ,Orch pods , 5. backup file : showtech_2024.02.20_08.56.45.tar\showtech_2024.02.20_08.56.45\mariad 6. Resource Usage CPU,Memory : showtech_2024.02.20_08.56.45.tar\showtech_2024.02.20_08.56.45\resource_usage All node with node utilization : describe_node Top nodes utilization Top pods utilization When telemetry is not accessible Check : kubectl get pods -A commonns | grep vm-grafana -- > pod is running [INST]: 21.1 All Callhome connections report down - known issue [/INST]: Root cause: RULES TO RECTIFY CALLHOME BULK FAILURES: * For systems with less than 3K connectors please use the MOP ESMBD-31557 * For systems with more than 3K connectors please use this MOP INSTRUCTIONS TO CLEAR BULK CALLHOME ISSUES WITH SYSTEMS > 3K CONNECTORS: 1. Pre-requisite: Verify this is not caused by an issue with the SGW. 2. Pre-requisite: Log into Controller and run the command: netstat -anp 4334 | grep EST | wc -l Expectation: It should match the current # of connector established connections if not, then run the following command: netstat -anp 4334 | grep SYN | wc -l netstat -anp 4334 | grep CLO | wn -l PROCEDURE : 1) stop the controller services in non vip node followed by vip node sudo systemctl stop controller (i.e on all nodes) //wait for controller to stop, check using "sudo systemctl status controller", sudo systemctl status controller (i.e on all nodes) 2) After stopping controller service on all nodes run this on all the 3 nodes (i.e global rule to block callhome) sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 -j DROP sudo iptables -L Must be done: Make sure the source port range is checked, as it could deviate from the below range Add below rules on 3 nodes sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 3000:5000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 5000:10000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 10000:15000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 15000:20000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 20000:25000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 25000:30000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 30000:35000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 35000:40000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 40000:45000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 45000:50000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 50000:55000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 55000:60000 -j DROP sudo iptables -A INPUT -i eth0 -p tcp --destination-port 4334 --source-port 60000:63000 -j DROP Run "sudo Iptables -S" 3)start all 3 controller instances in parallel sudo systemctl start controller (wait for some time) //wait for controller to start check using "sudo systemctl status controller", //wait for 2 min – this is necessary sudo systemctl status controller (Ensure controller service is running) Make sure all nodes are up: grep -ir "aye captain" /opt/controller/data/log/karaf.log --> time to see this line in log may vary --> Have patience!! 4) Check node and cluster status on all 3 nodes nodeStatus clusterStatus.sh It should be showing the appropriate output sudo iptables -L ----> to list iptables rule [INST] - Troubleshoot Director health Status from showtech [/INST] : Go to : Granite\Director_20230905_193122_Rel.19.1.1_P2_31\Director_20230905_193122_Rel.19.1.1_P2_31\Director_20230905_193122\APPLICATION\Dir_Application_LOG.tar\Dir_Application_LOG\var\opt\adva\log\aos\nc\director-cluster.log * This Log will have traces every 10 secs overall director node Status : * If its Healthy : INFO | VM Monitoring | Director: OK * Not Healthy : Director: FAIL, VM Monitor: INTERNAL. * I look out for Director FAIL always in this log, & No of occurrences , if 3 consecutive failure, Director VM would reboot Line 28634: 2023-08-27 10:10:41 +00:00 | INFO | VM Monitoring | Restarting vm after detecting process failure. Line 204938: 2023-09-05 18:35:51 +00:00 | INFO | VM Monitoring | Restarting vm 1) How to confirm Director Reboot history from director-cluster.log Grep the log for the following signature "load average: 2)How to confirm the "manual restart of director services" signature from the director-cluster.log You can grep for this signature "Starting Director Services [manual start]" in the director-cluster.log,if you see the match like below it indicates there is a manual restart of director services. 2023-10-04 11:03:36 +00:00 | INFO | DirectorStart | Starting Director Services [manual start] ... 3)How to confirm the "OS restart director services" signature from the director-cluster.log You can grep for this signature "Starting Director Services [OS Startup]" in the director-cluster.log, if you see the match like below it indicates there is a Automatic restart of director services. 2023-10-04 11:04:09 +00:00 | INFO | DirectorStart | Starting Director Services [OS Startup] ... 4) Check in secgw master and backup node for recent switchovers with these signatures journalctl | grep Entering journalctl | grep -i Entering 5.Call home disconnected signature in controller acorss all karaf logs (open mobaxterm and navigate to controller karaf log location) Karaf log location in controller showtech:Controller_20231229_070729\CONTROLLER\log grep -i "" kara* > CallhomeDisconnected.txt Ex. karaf.log:2023-12-29T05:22:53,218 | INFO | DOMNotificationRouter-disruptor-0 | ControllerConfigCtlrEventListener | 72 - com.adva.ensemble.controller.controllerconfiguration-impl - 21.1.1.6 | Successfully executed the release-auth-token RPC. Response Code: 200, Msg: If you noticed a significant no of call home disconnects for large no of connectors within a short period of time followed by a sec gw switchover you can confirm the issue. [INST]: LDAPs enabling by certificate type [/INST] : Here are steps we need to perform while enabling secure ldap communication at Lowes. There are two parameters that needs to be changed/edited. * Change the port from normal to secure port (636) available as part of peer system. * Upload the public certificate file as part of LDAP peer system. * Self Signed and CA signed Certificate: * Upload public key to peer system. * Perform script execution the script name will be displayed while saving the peer system. * Login into Director * type in "alias". A list of scripts will be displayed * Select "update_ldaps_certificate" * Internal CA Signed: * Upload public key to peer system. * Perform script execution the script name will be displayed while saving the peer system. * Login into Director * type in "alias". A list of scripts will be displayed * Select "update_ldaps_certificate" * Upload CA certificate to all director and execute below command. keytool --import -alias cerldapsha1 -file <> -storepass changeit -keystore "/usr/lib/jvm/java-8-openjdk-amd64/jre/lib/security/cacerts" [INST]: DCHP option 148 patch instructions [/INST] : This ticket contains Pre-requisites and instructions to apply the patch for Etisalat customers that require such option. There should be a TAC ticket to track the device mms hostname until feature is delivered in 3.6.x GA [INST]: MANO Install failure /Rectification : Orch Galera Formation Failure Signature. Mariadb Fails to start up [/INST] : MANO Restore log path on director node where mano restore & start scripts are executed /var/opt/adva/log/aos/nc Files mano_restore.log mano_console.log Director Restore Log file path /var/opt/adva/log/aos/nc/restore.log Director Cluster log path & file /var/opt/adva/log/aos/nc/restore.log Controller restore,cluster log file path /opt/controller/data/log restore.log cluster_configuration.log Orch restore , cluster log file path /var/log/ensemble/eso-core/restore.log /etc/opt/ensemble/cluster.log [INST] - Fix controller process start impacting callhome reestablishments [/INST] : IMPACT – TAC Observations on Lumen setup Case - 00305812,00307542 ESMBD-39619 - TAC - All connectors call home Down post patch 20.2.1.5 on director UI Release 20.2.1.5 or 20.2.1.4 Callhome doesnt establish after controller process is started or controller is rebooted. netstat -anp | grep 4334 | grep CLOSE --- o/p would result huge no of Sessions in Close wait than the provisioned Connectors. PROCEDURE TO APPLY THE MOP: Login to controller each node in cluster/Standalone as ensemble user and perform below step. 1. Take backup of the start script Cluster each node ,take the backup - cp /opt/controller/bin/start ~/start_bkp1 Standalone - cp /opt/controller/bin/start ~/start_bkp1 2. Copy the attached script ( “start” ) to /opt/controller/bin/ path as ensemble user in each node. No need to perform any restart after copy, this script will take into effect in the next controller service start. 3. Check sudo systemctl status controller 4. Check nodeStatus should report all Ok 5. Lumen team to monitor callhome Issue post controller restart(if any). [INST]: How to troubleshoot PPPoE link down issues [/INST] : How to troubleshoot PPPoE connectivity in case it goes down. 1. Take a troubleshooting report and open the journal.txt You can see the issue in the log by the fact that we stop to restart the ppp-client if the connection fails. In the log when the ppp-client is not able to established the connection you have: Mar 24 13:00:55 DI-1445-ST1-vRouter-onlyvr pppd[27287]: Remote message: Access number is exceedMar 24 13:00:55 DI-1445-ST1-vRouter-onlyvr pppd[27287]: PAP authentication failedMar 24 13:00:55 DI-1445-ST1-vRouter-onlyvr pppd[27287]: Modem hangup Mar 24 13:00:55 DI-1445-ST1-vRouter-onlyvr pppd[27287]: Connection terminated.Mar 24 13:00:55 DI-1445-ST1-vRouter-onlyvr pppd[27287]: Exit. 1. Before you try to solve this issue, as a root user you can do the following operations before trying to solve the issue via nc-cli or by systemctl: # dpkg -i python-pip-whl_9.0.1-2_all.deb # dpkg -i python3-pip_9.0.1-2_all.deb # python3 -m pip install py_spy-0.3.11-py2.py3-none-manylinux_2_5_x86_64.manylinux1_x86_64.whl -f ./ --no-index --no-deps The deb and whl files are attached in ticket 15558. The packages should be copied under root directory and you must run the parameters from the same directory folder that "whl" is installed. # py-spy dump --pid -> you can get the pi of yams with 'ps aux | grep yams | grep python' ** packages are attached. 2. You can try to resolve this issue by running the commands: nc-cli edit running vrf interface pppoe enabled false commit vrf interface pppoe enabled true commit 4. If it does not work then you can try , as root: # systemctl restart ppp-client@.service and check that the service is correctly started (with systemctl status ppp-client@ [INST]: Truncated troubleshooting report [/INST] : ADVA observed that some of the Etisalat servers that reported issues with conntracks had truncated troubleshooting reports. If that happens, please request for the customer who is running a vrouter or se-gw to install the "troubleshooting-report.sh" script attached to this [INST] under /usr/bin/ folder. You must be logged as a root user. That is the script that is executed when you do a troubleshooting report This new version creates a file call journal-last-24h.txt that contains the content of journalctl of the last 24h so we will always have the last logs [INST] - 6wind Connections transmissions resulting on Packet Loss (Conntracks) [/INST] : This is for Etisalat scalability issues that hit some of their end customers in 2 hotel chains. This is a procedure to address Conntracks hitting its max value of 8152 (2GB memory) by default. Instructions below are to modify the following configuration parameters: Instructions attached on how to make the necessary changes for customer who hit this issue. How to troubleshoot this issue: 1. Collect troubleshooting report 2. Verify the file "fp_cli_filling.tx" and check parameter 'nfct (number of flows for conntrack). Example: 8192/8192 then it has reached its limits. 3. Journal.txt contains messages as: fpm_nfct_add: error table is full (8192), unable to add a new entry kernel: nf_conntrack: nf_conntrack: table full, dropping packet libfps: libfpn_shmem: mmap failed in hugetlb_shm_mmap(fpn-port-shared): Cannot allocate memory (memory leak) 4. Check cpu and number of flows as a root user cpu usage: show fast-path cpu-usage fp-cpu-usage conntrack -C fp-cli nfct4 fp-cli-nfct4 fp-cli nfct4 | grep "Number of flows" * Make sure to apply the MOP Performance file recommendations with respect to the fast path parameters. filter4-max-ct 80000 [INST] : Revert a VM from failed to running (Newton) [/INST] : 1) Gain privileged access to the Connector (i.e. ovnchallenge) Pre-requisite: Verify that Vm is down - "show vnf info" 2) Enter the Openstack controller '/ovn/bin/dockersh ovn_ctrl' 3) Obtain the UUID of the server which is in the error state 'openstack server list' 4) Reset the VNFs state back to Active 'nova --insecure reset-state --active ' 5) Verify in the output from the command in #4 that the reset was successful. (could look at the server list again as well. 'openstack server list' VM should be in the Active state in the DB (though not really running yet) 6) Hard reboot the instance 'openstack server reboot --hard ' 7) Verify VM is up and running from CLI command "show vnf info" Note: If VM is not up, MANO might report "green" if the VM is active but not in "running state". Then investigate why VM is not coming up fully. It could be other issues such as "libvirt issue" a restart of "libvirt" wait a few minutes and do a hard reset from Orchestrator. If still does not work, then as last resort do a "reboot" from Connector to unblock the customer. 8) Orchestrator should report VM up and running [INST] - Script to CleanUp Locked Session in 19.1.1 [/INST] : ESMBD-30868 - TAC- Cannot Delete the device from ADVA Director Procedure to detect the Problematic Locks & cleanup the locked Up session DB command to check the transaction lock issue. installer@app1:~$ mongo nc -u aos -p MongoDB shell version v4.0.9 Enter password: connecting to: mongodb://127.0.0.1:27017/nc?gssapiServiceName=mongodb Implicit session: session { "id" : UUID("b22b3f10-e2ed-45ef-8f76-dd09cdaab863") } MongoDB server version: 4.0.9 2020-12-21T15:29:38.461+0000 E - [main] Error loading history file: FileOpenFailed: Unable to fopen() file /home/installer/.dbshell: Unknown error rs:PRIMARY> db.getCollection('sec_ses').find({"trx_id":{$ne : null}}) rs:PRIMARY> [INST] - NTP Sync On Orch Nodes using ntp conf & dsiable chronyd [/INST] : MOP - NTP Sync using NTP conf On orchestrator - we have Chronyd & ntpd service options in Orchestrator, by default Chronyd service is enabled. Recent observations from Etisalat & TAC systems , it is noticed that NTP sync doesn't work or come up well as expected even after following Documentation. Follow below recommendations 1. Disable chronyd service before ntpd is started sudo systemctl disable chronyd 2. Follow NTP sync up procedure per Director Install Guide [INST] - Fix Common name on all controller nodes post restore action of 20.2 backup on 21.1.1 Systems[] [/INST] : [MOP -Fix Common name on all controller nodes post restore action of 20.2 backup on 21.1.1 Systems[] Problem - Sometimes common name are not restored on Controller Nodes , if common name has additional "-" etc. It is fixed in 22.0 Rel Workaround: After you upgrade from 20.2 to 21.1, manually modify the properties in the com.adva.ensemble.controller.cfg file on all controller node Below Procedure to be followed in Backup Stage of 20.2 systems 1. Backup the config file from 20.2 Controller System - com.adva.ensemble.controller.cfg file 2. Post 21.1.1 Restore action is completed successfully 3. Following changes to be done on - Login to all controller cluster nodes - Edit the file /opt/controller/etc/com.adva.ensemble.controller.cfg & apply changes on all 3 nodes common-name=advaoptical.com ---->>>> change the value per 20.2 backed up file. - sudo systemctl stop controller on all 3 nodes - sudo systemctl start controller on all 3 nodes 4. Later verify if it is properly updated [ensemble@controller ~]$ grep -i "common" /opt/controller/etc/com.adva.ensemble.controller.cfg common-name= [INST] - Update ZTP status [/INST] : ESMBD-40287 - TAC - Connector Stuck in Orchestrator Staging stage but ZTP completed status in Splash Screen > 48 Hrs Problem Statement: Connector ztp status Stuck in Orchestrator Staging state, but ZTP completed and service is operational. Service orch status remains in grey. Follow the below procedure for the same. 1. Login to director ssh terminal using installer user. In case of Cluster identify the Primary DB by cluster status command and login. 2. Connect to mongodb terminal using below command. Use database nc. sudo /opt/adva/nc/mongodb/bin/mongo admin -u root -p 4netMGMT use nc 3. Get the impacted connector ztp status using using below command. Change the ctr.uid accordingly and confirm that you are updating the right connector. 4. Update the Service Orchestration Status by running below command. This command will update the ztp status. Note,Please chaneg the ctr.uid to the uid of the connector to be updated. db.me_me.updateOne({ "ctr.uid" : "test1" }, { $set: { "ctr.cldepsts" : "[{\"stage\":\"ZTP_STATUS\",\"state\":\"success\",\"percentage\":100,\"timestamp\":\"2022-11-08T11:19:38.607Z\",\"information\":\"ZTP is success, commissioning-status:commissioning success, recent-update:Service Orchestration is successful\",\"order_no\":0,\"display-name\":\"FULLY UP\"},{\"stage\":\"AUTHENTICATION\",\"state\":\"success\",\"percentage\":100,\"timestamp\":\"2022-11-08T10:55:11.331Z\",\"information\":\"Authentication is successful\",\"order_no\":1,\"display-name\":\"Authentication\"},{\"stage\":\"CONNECTOR_CALLHOME\",\"state\":\"success\",\"percentage\":100,\"timestamp\":\"2022-11-08T10:55:54.892Z\",\"information\":\"Connector Call home is successful\",\"order_no\":2,\"display-name\":\"Connector Call Home\"},{\"stage\":\"CONNECTOR_DISCOVERY\",\"state\":\"success\",\"percentage\":100,\"timestamp\":\"2022-11-08T10:56:06.090Z\",\"information\":\"Commissioning is successful\",\"order_no\":3,\"display-name\":\"Connector Discovery\"},{\"stage\":\"EMBEDDED_CLOUD_READY\",\"state\":\"success\",\"percentage\":100,\"timestamp\":\"2022-11-08T11:12:37.523Z\",\"information\":\"Embedded Cloud is ready\",\"order_no\":4,\"display-name\":\"Embedded Cloud Ready\"},{\"stage\":\"SERVICE_ORCHESTRATION\",\"state\":\"success\",\"percentage\":100,\"timestamp\":\"2022-11-08T11:19:38.607Z\",\"information\":\"Service Orchestration is successful\",\"order_no\":5,\"display-name\":\"Service Orchestration\"}]" } }) [INST] - Log Rotation on Orch Syslogs & eso-networkctrl logs fine tuning [/INST] : Lumen complained of showTech collection is taking longer time >= 45mins 1. Check the disk space on the system - du -h 2. Check the file/folder sizes in /var/log/ path 3. Check log4j.xml if any debug is enabled for any of the process If any log is present >= XXGb , copying & compressing would definitely add time. Hence these factors add on time. Two such log were present - /var/log/messages , messagesXX,messages XY. [INST] - To clean up the KahaDB files in Orchestrator [/INST] : This MOP is to be executed on MANO systems running 18.4.1 & 19.1.1 build and when JMS failures are reported in Orchestrator. Following Conditions to be met for Executing MOP 1. Continuous traces of below lines must seen in orchestrator logs (BrokerService.java:581) - Failed to start Apache ActiveMQ (localhost, null). Reason: java.io.IOException: Detected missing/corrupt journal files. 3720 messages affected.java.io.IOException: Detected missing/corrupt journal files. 3720 messages affected. 2. Orchestrator Version in 18.4.1.p1,19.1.1 Up & running for longer time 3. Unable to see any visible logs/Errors while performing Deletion/Creation on UI/over logs Execute the following steps- 1. Delete all the files under /var/opt/ensemble/eso-core/activemq-data/localhost/KahaDB 2. Restart the eso-core service - “sudo service eso-core restart”. [INST] - To block vulnerable port 15672 from outside access in director instances [/INST] : This MOP needs to be executed to handle the security issues in RabbitMQ service in Director in 18.4.1 build. These security issues have been fixed in 19.1.1 build. Port 15672 needs to be blocked from outside access in director instances by default since Rabbit MQ admin http server running on this port has several vulnerabilities. * Login as "installer" user to Director VM * Take a backup of /etc/rc.local for safekeeping sudo cp /etc/rc.local /opt/rc.local_ORIG_BK * Configure Ip Table rule to block 15672 and make it persistent sudo iptables -A INPUT -p tcp --destination-port 15672 -j DROP sudo touch /etc/iptables.rules sudo chmod 666 /etc/iptables.rule sudo /sbin/iptables-save > /etc/iptables.rules sudo sed -i '/^#/!s/exit 0/sudo \/sbin\/iptables-restore < \/etc\/iptables.rules\nexit 0/g' /etc/rc.local see below the updated Procedure : Login as "installer" user to Director VM Take a backup of /etc/rc.local for safekeeping =============================================== sudo cp /etc/rc.local /opt/rc.local_ORIG_BK Configure Ip Table rule to block 15672 and make it persistent ============================================================= sudo iptables -A INPUT -p tcp --destination-port 15672 -j DROP sudo touch /etc/iptables.rules sudo chmod 666 /etc/iptables.rules sudo /sbin/iptables-save > /etc/iptables.rules sudo sed -i '/^#/!s/exit 0/sudo \/sbin\/iptables-restore < \/etc\/iptables.rules\nexit 0/g' /etc/rc.local [INST] : Access Keycloak DB [/INST] : Commands to access the keycloak server Login to director vip node and perform the following steps: * command to see the iptable rules: sudo iptables -L --line-numbers root@app1:/home/installer# sudo iptables -L --line-numbers Chain INPUT (policy ACCEPT) num target prot opt source destination 1 ACCEPT tcp -- anywhere 10.64.107.188 tcp dpt:http state ESTABLISHED 2 ACCEPT tcp -- anywhere 10.64.107.188 tcp dpt:https state ESTABLISHED 3 ACCEPT tcp -- 10.64.107.198 10.64.107.198 tcp dpt:http state ESTABLISHED 4 ACCEPT tcp -- 10.64.107.198 10.64.107.198 tcp dpt:https state ESTABLISHED 5 ACCEPT tcp -- 10.64.107.187 10.64.107.198 tcp dpt:http state ESTABLISHED 6 ACCEPT tcp -- 10.64.107.187 10.64.107.198 tcp dpt:https state ESTABLISHED 7 ACCEPT tcp -- 10.64.107.199 10.64.107.198 tcp dpt:http state ESTABLISHED 8 ACCEPT tcp -- 10.64.107.199 10.64.107.198 tcp dpt:https state ESTABLISHED 9 DROP tcp -- anywhere 10.64.107.198 tcp dpt:http state ESTABLISHED 10 DROP tcp -- anywhere 10.64.107.198 tcp dpt:https state ESTABLISHED 11 DROP tcp -- anywhere anywhere tcp dpt:15672 12 DROP tcp -- anywhere anywhere tcp dpt:8002 13 DROP tcp -- anywhere anywhere tcp dpt:8012 14 DROP tcp -- anywhere anywhere tcp dpt:8443 15 DROP tcp -- anywhere anywhere tcp dpt:http-alt ================this is to be removed *Command to remove or drop the tcp http-alt sudo iptables -D INPUT 15=========[15 is the linenumber from the last command] *command to save the iptable rules sudo iptables -S [INST]- LDAP Password change from Director UI Problem statement: User is not able to login to director GUI after changing the LDAP Password in director admin login [/INST] : LDAP Passwrod change Prerequisites: 1. LDAP password should be changed in LDAP server 2. Login to director UI with admin access and modify the LDAP password under peers 3. Check the connectivity is green for all the peer system Solution: Need to perform live debugging with engineering team Duration: 1 hour to 1:30 minutes Impact: 1. Director tenant login might be impacted during debugging 2. Director tenant- creation of preconfig for new customers/existing customers might be impacted Post session validation: 1. All the peer system connectivity should be green in director admin ui 2. LDAP authentication should be working fine and user should be login to director [INST] - Vrouter upgrade- 2 methods for DPI usecase [/INST] : Action: Vrouter in openstack should be upgraded from 3.4.0.ga to 3.4.11 Before upgrade copy the configuration. Method-1: 1. Login to orchestrator by doing ssh. 2. Login to vrouter from orchestrator by ssh root@<> 3. Download 6wind_x.x.update version from 6wind portal to a host 4. Copy running to startup VEC>copy running startup 5. Validate the configuration before upgrade VEC>validate startup 6. .Export backup configuration to host VEC>cmd backup export vrf data url scp://:@:22/pathTofile 7.Copy 6wind_x.x.update from host to 6wind Sec-GW VEC>cmd system-image import vrf datascp://:@:22/pathTofile<6wind-vrouter-tr-ae-x86_64-v3.1.2.update> Checking startup configuration compatibility with imported image ... 8. Check "new version is listed as next" VEC>cmd system-image list 3.1.0.ga (default) (current) 3.1.2 (next) 9. Set the latest version as default VEC>cmd system-image set-default 3.1.2 10. Reboot Sec-GW VEC>cmd reboot delay 0 11. After reboot under the root shell execute the below command to get the access to console from the connector : - systemctl enable serial-getty@ttyS1.service - systemctl start serial-getty@ttyS1.service [INST]- 6wind upgrade steps- 3.4.4m1 to 3.4.8 [/INST] : Below are the steps to perform upgrade: 1. Download 6wind_x.x.update version from 6wind portal to a host 2. Copy running to startup VEC>copy running startup 3. Validate the configuration before upgrade VEC>validate startup 4.Export backup configuration to host VEC>cmd backup export vrf url scp://:@/pathTofile 5.Copy 6wind_x.x.update from host to 6wind Sec-GW VEC>cmd system-image import vrf mano scp://:@:22/pathTofile<6wind-vrouter-tr-ae-x86_64-v.update> Checking startup configuration compatibility with imported image ... OK: startup configuration is compatible with version 6. Check "new version is listed as next" VEC>cmd system-image list 3.4.4.m1 (default) (current) 3.4.8 (next) 7. Set the latest version as default VEC>cmd system-image set-default 3.4.8 8. Reboot Sec-GW VEC>cmd reboot delay 0 9. After successful installation, Delete Old Version VEC>cmd system-image delete 3.4.4.m1 [INST]- Vrouter(tr-agent) upgrade - 2 methods for DI Usecase [/INST] : [MOP- Vrouter(tr-agent) upgrade 2 methods for DI Usecase Please note the below methods are applicable only for the vrouter which is installed as KVM in tr-agent container. This is not applicable for openstack vrouter installation This procedure upgrades the vRouter VNF package using Connector CLI commands. 1. Transfer image to Connector in the /tmp directory. 2. Use the command: userfile install vnf-update file:///tmp/ Example: userfile install vnf-update file:///tmp/6wind-vsr-x86_64-v3.4.11.update 3. Go to the Application-maintenance menu, and run: vnf-upgrade vrf "default" container tr-agent Example: vnf-upgrade vrf "default" container tr-agent 6wind-vsr-x86_64-v3.4.11.update [INST] - 6wind procedure to remove dsvti interface [/INST] : Please refer to the attached procedure. [INST] - 6wind upgrade procedure from NON HA pool to HA ike pool [/INST] : Here are the recommended steps to switch from a non HA pool to a HA pool configuration without service disruption. * On backup node, bring down physical interface through which HA sync is made ( identify the interface which is linked to vlan. example: vlan ens224.151). This isolate nodes from synchronization perspective * Save your configuration and copy running to startup * On backup, upgrade backup node with 3.6.b1.0.ga * On backup, after reboot change configuration in order to : * create HA pools you need, with DNS and subnet attributes. * delete former regular pool used in ike container * bind ha pool to vpn(s) * restore sync link by enabling again ens224 * check synchronization is done and that all IKE SAs are is passive state. * On master node, trigger a failover by disabling ens160 or ens192 * On master node wait end of failover ( IKE SAs in passive state), disable synchronization link by disabling ens224 interface. * Save your configuration and copy running to startup. * On master node upgrade to v3.6.b1.0.ga * On master node, after reboot change configuration in order to : * create HA pools you need, with DNS and subnet attributes * delete former regular pool used in ike container * bind ha pool to vpn(s) * restore sync link by enabling again ens224 * check synchronization is done and that all IKE SAs are is passive state. * On master node save configuration and copy running to startup * On master node, check synchronization is achieved again ( IKE SAs in passive state) * On master node, enable again interface ens160 or ens192 to trigger a new failover and switch back to last upgraded node. * Check IKE SAs go to Establish state. Check IKE SAs are in passive state on peer. * On master node save configuration. If you are satisfied with setup then define last version as the default one for boot. [INST] - Non-active vports Cleanup ,if No of Interfaces(gig,vports - Active+Non-active) > 50 limit & Process SysRepoSync fails to come up [/INST] : (image: /polarion/icons/default/enums/type_bug.gif) ESMBD-32913 - TAC - Connector reports sysrepo and health check failures | * SysRepoSync Code Limit - in 18.4.1. & 19.1.1 Connectors : SysRepoSync Code login can hold upto 50 interfaces(Oper Up + Oper Down)entries only. * Impact – Since SysRepoSync Code queries no of interfaces in a cyclic manner & due to limit of 50 within Code. It Continuously fails & clears & hence fails to restore Callhome .This happens in a cyclic fashion. Ongoing (Clr/fail) Process SysRepoSync alarms are visible & can be noticed on Cli. * Best Practices untill Fix available – * Plan well NST in advance on No of vPorts. * Define only required No of vports in VNF Definition within <=50(count should include vports + gig /tengig interfaces)in Orchestrator. * Do not over configure Trailing ports in VNF which would’t be used in future. * Workaround: * If Process SysrepoCode Alarms are continuously failing/clearing in 18.4.1/19.1.1 * Identify No of interaces : show interfaces , Check if it exceeds 50 (Oper Up + Oper Down) * Disblae netconf on connector config>netconf netconf>no enable netconf>commit [INST] - Service Edit failure in Delete VNF Flow [/INST] : MOP - Service Edit failed due to unsuccessful VNF boot up This MOP would be applicable for service edits which involves Delete VNF steps.Kindly confirm with customer about service edit steps 1. During Service Edit 2. Service Edit flow, Final Attempt VNF state shown as unknown 3. Service rollback fails due toException at stop VNF state .Hence it marks as Service as Failed 4. Find out from the Orchestrator log @ which step did the service edit fail 5. Take the mysql dump from the orchestartor terminal sudo mysqldump -u ensemble -p osa > backup.sql 6. Login to DB & Verify Rollback Step for the failed SERVICE, Use service name. mysql -u ensemble -p use osa select * from service_order_delta where service_order_id =(select id from service_order_desc where name='USA-SJ-P01-POD02-C03-SDWANGW'); 7. Update the rollback step on the failed service. select id from service_order_desc where name='USA-SJ-P01-POD02-C03-SDWANGW'; update service_order_delta set rollback_step="decomissionVNFinstancesForDelta' where service_order_id 860; 8. Verify the service select * from service_order_delta where service_order_id = 860; 9. Update the service status to FailedToEdit, by doing this it enables Manual rollback option on service row details in Orch UI. update service_order_desc set status = 13 where id=860; 10. After above step,it enables Manual rollback option on service row details in Orch UI and Service Status marked as FailedToEdit. 11. Click on Rollback, Verify the Service states reverted to Original state before edit transitions from FailedToEdit --> Pending ---> Running (Orginial state before Edit) 12. Verify on connector , problematic VNFs are cleaned up Successfully. [INST] - Service Edit failure in Add VNF Flow [/INST] : MOP- Service Edit failed due to unsuccessful VNF boot up This MOP would be applicable for service edits which involves ADD VNF steps.Kindly confirm with customer 1. During Service Edit,new VNF booting taking longer time, Even after 180 tries. Instantiating VNF SDWGW0000007777SNG-MCP of boot order 1 2. Service Edit flow, Final Attempt VNF state shown as unknown Retry attempt : 3 VM_Status : UNKNOWN , VNF_Name : SDWGW0000007777SNG-MCP. 3. Service rollback fails due to unknown VNF state .Hence it marks as Service as Failed Cannot Proceed further as the VNF state is not ready yet. Service status set to failed as we cannot revert the service back to original Service status set to failed 4. Find out from the Orchestrator log @ which step did the service edit fail 5. Take the mysql dump from the orchestartor terminal sudo mysqldump -u ensemble -p osa > backup.sql 6. Login to DB & Verify Rollback Step for the failed SERVICE, Use service name. mysql -u ensemble -p use osa select * from service_order_delta where service_order_id =(select id from service_order_desc where name='USA-SJ-P01-POD02-C03-SDWANGW'); 7. Update the rollback step on the failed service. select id from service_order_desc where name='USA-SJ-P01-POD02-C03-SDWANGW'; update service_order_delta set rollback_step="undoConfiguredVNFs" where service_order_id 860; 8. Verify the service select * from service_order_delta where service_order_id = 860; 9. Update the service status to FailedToEdit, by doing this it enables Manual rollback option on service row details in Orch UI. update service_order_desc set status = 13 where id=860; 10. After above step,it enables Manual rollback option on service row details in Orch UI and Service Status marked as FailedToEdit. 11. Click on Rollback, Verify the Service states reverted to Original state before edit transitions from FailedToEdit --> Pending ---> Running (Orginial state before Edit) 12. Verify on connector , problematic VNFs are cleaned up Successfully. [INST]: Deletion of devices gets stuck on "deleting", "adding" or "creating" state or Remove service in FailedtoDelete [/INST] : Tracks [INST] deliverable, procedural steps and updated script. In case deletion of service only from Orch UI - FAILEDTODELETE without Cloud reset decision By customer 1. Restore connector connectivity 2. Reattempt delete on Service with (FailedToDelete) If delete fails again in Scenario : , Involve Engg for assistance to remove service entity [INST] - Granite - Collect Controller Stats [/INST] : Controller – Performance Stats copy monitor_controller.sh to /home/ensemble· add cron jobcrontab -eAdd below entry & save the list*/1 * * * * /home/ensemble/monitor_controller.sh· Check if entry existscrontab -l Script collects network stats & IO stats & controller process stats. Data helps Engg team to debug if there any really network issues around issues reported. [INST] - Update eo-poller-iterations to 100 in 19.1.1 [/INST] : ESMBD-30953 - TAC - ZTP is failed, commissioning-status:commissioning success, recent-update:Service Orchestration has failed as creation poller reached maxRetries 30 Procedure to detect the Problematic Locks & cleanup the locked Up session Details - controller stop update the parameter in the cfg file controller start From the analysis done, looks like this is a bug in the controller code. The value of 1080 iterations is not being honored for service orchestration. This will be addressed in 20.1, in the current sprint itself As a workaround, the customer can set the value of eo-poller-iterations to 100. This will give the service orchestration step 100 minutes to complete. Use the above mentioned steps to udpate teh value in cfg file. Steps : - 1. sudo systemctl stop controller 2. Edit the file - vi /opt/controller/etc/com.adva.ensemble.controller.cfg 3. update eo-poller-iterations=100 & save the file 4. sudo systemctl start controller [INST] - If customer decides to migrate existing tenants from Local Auth/LDAPs to SAML Authentication [/INST] : * Disable remote_authenication flag in Ensemble_auth.conf. * Enable(True) SAML authentication flag in Ensemble_auth.conf. * Create SAML peer system for domain and as well as for tenant SAML authentication.(Note: Follow the user guide for the SAML peer system configuration * Associate SAML tenant peer system configuration to tenant. * Make sure LDAP email address matches with SAML user email address for the same account.(Note: It is up to IdP provider to make sure the LDAP users exists in IdP and associate the users to the SSO application.) * Make sure LDAP email address matches with SAML user email address for the same account. * Perform aos_restart. [INST] - Script to Delete Connectors in director , Only when Deletion State is hung in "Delete In Progress" State > than Expected Time [/INST] : ESMBD -32070 TAC - Connectors Stuck in delete In progress State > 8hours in Director Procedure to detect the Problematic State in Director System 1. User invoked Delete Connector in Director 2. Controller Event is successfully received "Connector pre-configuration deletion successful in controller" ,Check in app1_app-tier-scala-container.log 3. After event is seen, Check director log - app1_app-tier-scala-container.log for below trace/print. >>"Unable to delete Connector" , DeleteCIMResponse(1.0,Status(404,List(Detail(0,9,List(),,))),Map(),rsp,dlt). Retrying... Execution Steps. 1. Copy the script to /home/installer path in director system 2. Provide the executable permission : chmod 777 3. Execute the script - Director_DeleteConnectorInprogress_hung.sh TAC: 21.1 All Callhome connections report down - Call Home Service Under nodeStatus "Not OK" [/INST] : NOTE:This Procedure should be only used With Engg Confirmation,DO NOT TRY ON YOUR OWN .This is not a normal callhome MOP For Normal Call Home [INST] Kindly refer ESMBD-45778 RULES TO RECTIFY CALLHOME BULK FAILURES: * For systems with less than 3K connectors please use the MOP ESMBD-31557 * For systems with more than 3K connectors please use this MOP INSTRUCTIONS TO CLEAR BULK CALLHOME ISSUES WITH SYSTEMS > 3K CONNECTORS: 1. Pre-requisite: Verify this is not caused by an issue with the SGW. 2. Pre-requisite: Log into Controller and run the command: netstat -anp 4334 | grep EST | wc -l [INST] - Unable to find Mac address [/INST] : :grep -i "No MAC addr file for port" /var/log/iso-install.log : grep -i "No MAC addr file for port" /var/log/dmesg : grep -i "selected as a suitable match" /var/log/iso-install.log : Comparison maestrOS.conf PCI mapping Vs dmesg PCI mapping - Customer confirmed Dell SFP+ installed on port 7/0.6 - Only 7 ports detected in BIOS - Config was tried with 8 ports - Customer reinstalled OS , it worked. So it means in such a case - Customer definitely to try mutiple attempts Installations. Singtel to confirm what was changed from HW side before last success. As adva stated no Issue in ADVA OS , there is a issue in way - it is setup via BIOS - manual config change in maestrOS - HW less resource than configured in OS maestrOS conf or cli configuration [INST] - Controller karaf log analysis [/INST] : Below are the different log signatures for different issues. Pick up as needed Path to Log on system live : cd /opt/controller/data/log showtech log Path : Controller.147_20240123_203459.tar/Controller.147_20240123_203459/Controller_20240123_203459/CONTROLLER/log 1. To check if controller process was stopped , this happens when there is controller stop gracefully,abrupt controller process restarting due to internode communication failures due to n/w delay grep -i "shutdown command. Stopping framework" karaf.log* or Controller started grep -i "org.apache.karaf.main.lock.SimpleFileLock lock" karaf.log* 2. To check Call home Issue on all nodes grep -i "Too Many Open files" karaf.log* ---> if above signature is found in current karaf.log , then call home MOP is highly required 3. To check Network delays which can cause internode communication grep -i "scheduled sending of heartbeat was delayed. Previous heartbeat was sent" 4. To check VIP fail overs grep -i " Ownership is updated with" 5. Node Quarantined post Network glitch/failure/split brain state grep -e "Got quarantined by" karaf.log 6. Call Home Disconnected grep -i "Callhome Name:CallhomeDisconnected Result:0" karaf.log 7. Call Home connected grep -i "CallHome request received" karaf.log 8. Single command to fetch all sort of logs in a sequence. grep -e "Got quarantined by \|org.apache.karaf.main.lock.SimpleFileLock lock\|received shutdown command. Stopping framework" * | cut -d":" -f2,3,4,5,6 | sort 9. check UID is present in Controller Datastore if nothing found in logs, sometimes NEC shares incorrect showTech CDS datastore : Etisalat\Controller_20240227_065557\Controller_20240227_065557\CONTROLLER\datastoreCallHome request receivedCallHome request receivedCallHome request received [INST] - Service Edit failure in Delete VNF Flow [/INST] : Steps for Recovery: 1)Login to the orchestrator UI to check the service status, the service is showing in a failed state. 2)Check the cluster status output from the orchestrator VIP node it should be showing proper output. (i.e all process and nodes should be showing operational) 3)Check the connector for ceilometer,disk space and rest lock up issue by using the MOP's previously shared to customer. show status df -h telnet 5000 The Above commands will help us to check the ceilometer,disk space issue and rest lock up issue. 4)Take the backup of the orchestrator DB from the VIP node using the following command 5) Login to the Orchestrator DB i) mysql -u ensemble -p NOTE:Default password is ensemble ii)use osa 6) Execute the following DB commands on the orchestrator to remove the vnf from the service. MariaDB [osa]> select id, name from service_order_desc where name like '%CHN-GZ-P01-POD02-C04%'; MariaDB [osa]> select id, name, virtual_machine_id, device_type from device where id in (select device_id from service_order_device where service_order_id=2233); MariaDB [osa]> select id from service_order_change where port_id in (select id from port where virtual_machine_id=3271); MariaDB [osa]> select id from port where virtual_machine_id=3271; MariaDB [osa]> select id, name, guid, service_order_id from virtual_machine where id = 3271; MariaDB [osa]> select id, cpus_granted from vnf_cloud_resources_granted where vnf_id=14800; MariaDB [osa]> select id, value, service_order_id from vnf_property where vnf_id=14800; MariaDB [osa]> select service_order_id, int_value, str_value , opt_type from optimization where vnf_id= 14800; MariaDB [osa]> select ep1_id from interconnection where service_order_id=2233; MariaDB [osa]> select id from endpoint where device_id=14800; MariaDB [osa]> select id from interconnection where service_order_id=2233 and ep2_id in (select id from endpoint where device_id=14800); MariaDB [osa]> select id from service_order_subnet where interconnection_id in (select id from interconnection where service_order_id=2233 and ep2_id in (select id from endpoint where device_id=14800)); MariaDB [osa]> select id, name from port where device_id=14800; MariaDB [osa]> MariaDB [osa]> delete from service_order_change where port_id in (select id from port where virtual_machine_id=3271); MariaDB [osa]> MariaDB [osa]> MariaDB [osa]> delete from vnf_cloud_resources_granted where vnf_id=14800; MariaDB [osa]> delete from vnf_property where vnf_id=14800; MariaDB [osa]> delete from optimization where vnf_id= 14800; MariaDB [osa]> delete from service_order_subnet where interconnection_id in (select id from interconnection where service_order_id=2233 and ep2_id in (select id from endpoint where device_id=14800)); MariaDB [osa]> delete from endpoint_info where interconnection_id in (select id from interconnection where service_order_id=2233 and ep2_id in (select id from endpoint where device_id=14800)); MariaDB [osa]> delete from service_order_change where interconnection_id in (select id from interconnection where service_order_id=2233 and ep2_id in (select id from endpoint where device_id=14800)); MariaDB [osa]> delete from interconnection where service_order_id=2233 and ep2_id in (select id from endpoint where device_id=14800); MariaDB [osa]> delete from endpoint where device_id=14800; MariaDB [osa]> delete from port where device_id=14800; MariaDB [osa]> delete from service_order_device where service_order_id=2233 and device_id=14800; MariaDB [osa]> delete from device where id=14800; MariaDB [osa]> delete from port where virtual_machine_id=3271; MariaDB [osa]> delete from virtual_machine where id = 3271; MariaDB [osa]> MariaDB [osa]> delete from service_order_delta where service_order_id=2233 7)Check the polling status MariaDB [osa]>desc cloud; MariaDB [osa]> select id,name,last_polled_at,resource_polling_state,now() from cloud where name like 'CHN-GZ-P01-POD02-C04'; -Check the vnf that we wanted to remove is removed from the orchestrator UI on the respective orchestrator service. -Wait for one more polling cycle 8)Check the Orchestrator Log(i.e on orchestrator VIP) status for the respective service,to check if the service is working or there were any issues on the service grep -i 'CHN-GZ-P01-POD02-C04-SDWANGW' orchestrator.log grep -i 'CHN-GZ-P01-POD02-C04-SDWANGW' orchestrator.* If the Logs were not seen(i.e on the (recent time stamps) go to step 9,10 9) Perform a Dummy service edit to add a network link to the existing service on to a new vnf without impacting DATA PATH Traffic and ensure Orchestrator logs for the service is being shown. 10) Check the service is showing the correct status and working properly from Orchestrator UI. [INST] : Director DB commands to unset the CIDR attribute in Tenant [/INST] : This specific commands to be run only when specific signatures are seen on UI Via admin login any serach tenant name using filter DNS NAME , NAME would appear with error : Attribute 'cidr' is not modelled for entity of type '/cim/mm/moc/mgmt,md'", "err": -1}] app1_C* logs will capture this signature, Signature appears after PATCH API call 2023-12-24T10:08:43.515Z|RESTAPIHandler|INFO|79bac4b816729c0d||Got PATCH call on /mit/md/655b0e2583ae2d0e18114a15 2023-12-24T10:08:44.061Z|RESTAPIHandler|INFO|79bac4b816729c0d||Connection closed 2023-12-24T10:08:44.218Z|RESTAPIHandler|INFO|c47141e0aae8913||Got GET call on /mit/md/655b0e2583ae2d0e18114a15 2023-12-24T10:08:44.251Z|RESTAPIHandler|INFO|c47141e0aae8913||Outbound error payload: {"rsn": "Bad Request", "dtl": [{"app": -1, "exp": "Attribute 'cidr' is not modelled for entity of type '/cim/mm/moc/mgmt,md'", "err": -1}]} To run below queries , Tenant ID should be known 1.Login to Database Primary 2. Switch to nc database rs:PRIMARY> use nc switched to db nc 3. Get Tenant array rs:PRIMARY> db.mgmt_md.find({'_id':'/mit/md/655b0e2583ae2d0e18114a15'}); 4. SOMEHOW CIDR key WAS SET as highlighted : 5. HENCE BELOW QUERY WAS USED TO UNSET CIDR key rs:PRIMARY> db.mgmt_md.update({'_id':'/mit/md/655b0e2583ae2d0e18114a15'}, {$unset: {cidr:1}}); WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 }) rs:PRIMARY> rs:PRIMARY> rs:PRIMARY> 6. Rerun step 3 to check CIDR is removed towards end rs:PRIMARY> db.mgmt_md.find({'_id':'/mit/md/655b0e2583ae2d0e18114a15'}); 7. Check if tenant logins are working & is able to list in super admin page . [INST] - Unable to delete Tenat due to assosciated templates exists [/INST] : ***THIS IS A SIDE EFFECT OF POST CALL HOME ISSUE. This can happen if the operations team is trying to delete a tenant as CONTROLLER and DIRECTOR are still synchronizing. Sequence ->First time creation will fail with Error "Ask timed out on [Actor[akka://NCActorSystem/user" on Director UI. -> Template doesn't appear in Director UI however in controller it would have succeeded -> It needs thorough log analysis -> if logs are not available,request controller showTech TAC Triage 1. Request Controller showTech 2. Untar showTech 3. Goto D:\Etisalat\Controller_20231226_144436\Controller_20231226_144436\CONTROLLER\datastore\ 4. Open cds.json in notepad++ >>>>Takes lot f time depending on file size.Have patience 5. Run the plugin -> format json 6. then cds.json will appear in json format 7. Look for tenant name in cds.json eg : RABIA INSURANCE CO S.L.A - 900104220188 8. Find all in cds.json using tenant id. search results should list all associated templates. in this case, only one will be dangling 9. Copy the template ID & use it in below curls [INST]: In Director GUI - Overall ZTP Status in grey state even though ZTP completed on connector [/INST] : Run cluster status on director node check healthy status login to dir node [INST] Orchestrator Polling Issue in 19.x.x due to eso-instance stopped polling impacted by kahadb signatures [/INST] : Orchestrator Polling Issue in 19.x.x due to eso-instance stopped polling impacted by kahadb signatures MOP to be applied on Orchestrator_cluster to resolve polling issue which leads to service edit failure. Eg VNF booting fails beyond 180 tries Steps to be followed: 1. Existing ZTP completed - While Inserting VNF trough service edit instance polling would report as not still booting. even when VNF is oper in connector 1. Login to all the 3 orchestrator via ssh 2. Run cluster_Status.sh on all the Orch cluster nodes 3. Access database on the Orch cluster nodes 4.Check within DB 5. sudo mysql 6. select * from eso_instances; o/p should return 3 entries . i.e 3 nodes 7.select id, name, last_polled_at, eso_instance_id from cloud where eso_instance_id; o/p would list N num of clouds with eso-instance id for which last poll update time would be never same as other clouds with different eso-instance id 8. Run stop_monitor command on all the Orch cluster nodes to disable VM monitor 9. Run systemctl stop eso-core command on eso-instance id for which polling had stopped working 10. Then start systemctl start eso-core 11.Check within DB on all the Orch cluster nodes sudo mysql select id, name, last_polled_at, eso_instance_id from cloud where eso_instance_id is NULL; select id, name, last_polled_at, eso_instance_id from cloud where eso_instance_id is not NULL; 9.Rerun the cluster status output on all the Orch cluster nodes cluster_Status.sh 10.Enable VM monitoring by running the following command on all the Orch cluster nodes to enable VM monitor 11.Access the Orch UI.Perform the Service Edit & check if Issue resolves. [INST]: [Director] Director UI does not reflecting the correct ZTP status [/INST] : Step1: Login to the director database primary Step 2:Log in to the director DB by typing the following command: installer@app1:~$ sudo /opt/adva/nc/mongodb/bin/mongo admin -u root -p 4netMGMT MongoDB shell version v4.0.9 connecting to: mongodb://127.0.0.1:27017/admin?gssapiServiceName=mongodb Implicit session: session { "id" : UUID("72ce685f-faab-4a92-90f0-e47c2cf1f836") } MongoDB server version: 4.0.9 rs:PRIMARY> use nc switched to db nc Step 3: Run the following get command to fetch the ZTP status of the UID from the DB rs:PRIMARY> db.getCollection("me_me").find({ "ctr.uid" : "900103876283-900103876279"}, { "mgmtip" : 1.0, "ctr.cldepsts" : 1 , "ctr.uid" : 1 } ); Step 4: Run the following command to update the correct ZTP status for the UID with time stamp on to the DB.: In the below command "900103876283-900103876279" is the uid and timestamp\":\"represents the ZTP time (i.e you can change the time values according to activation timestamp ) Step 5 : Go check in director UI , overall ztp status should reflect as green with all stages as Green & OPer up as green [INST] to Migrate Orchestrator VM's (in 2nd & 3rd cluster )from New Environment to Old Environment for Stability [/INST] : The Etisalat cloud team encountered that the new environment has some bugs related to network connectivity which makes the VM unreachable over the production network. Now recommendation from the cloud team is to move the VM(s) from the new to the old environment followed with power off the VM(s) and then moving to the old stable environment List of Vm's to be migrated From New Environment to old Environment: HRC-ND-ADVA-Prod-ORCHESTRATOR04 New Require migration HRC-ND-ADVA-Prod-ORCHESTRATOR05 New Require migration HRC-ND-ADVA-Prod-ORCHESTRATOR07 New Require migration HRC-ND-ADVA-Prod-ORCHESTRATOR08 New Require migration HRC-ND-ADVA-Prod-ORCHESTRATOR09 New Require migration Recommended Downtime for this Activity: 2 Hours Steps to migrate the VM's: Orchestrator Cluster-2 - VM-4, VM-5, VM-6 Keep the VIP as is. Move the standby nodes to the old env 1. Check which VM has the VIP assigned – clusterStatus 2. Currently VM-4, VM-5 is in new environment 3. If VM -6 is MAster then follow the below steps: a. Shut down VM-5 b. Migrate VM-5 to old environment c. Wait for the VM-5 to become operational. -- clusterStatus and nodeStatus d. Shut down VM-4 e. Migrate VM-4 to old environment f. Wait for the VM-4 to become operational. -- clusterStatus and nodeStatus 4. Check overall cluster Status Orchestrator Cluster-3 - VM-7, VM-8, VM-9 1. Check which VM has the VIP assigned -- clusterStatus (consider VM-7) 2. Keep VM-7 in new environment,Follow below steps: a. Migrate VM-8 b. Shut down VM-8 c. Migrate VM-8 to old environment d. Wait for the VM-8 to become operational. -- clusterStatus and nodeStatus e. Migrate VM-9 f. Shut Down VM-9 e. Migrate VM-9 to old environment f. Wait for the VM-9 to become operational -- clusterStatus and nodeStatus a. Shut Down VM-7 b. Migrate VM-7 to old environment c. Wait for the VM-7 to become operational -- clusterStatus and nodeStatus [INST]: To Remove Connector Peconfiguration Entries stuck at RESET-IN-PROGRESS in Director UI [/INST] : 1. Copy the script to /home/installer path in director system 2. Provide the executable permission : chmod 777 3. Execute the script - Director_DeleteConnectorInprogress_hung.sh Please make sure to use this script only when Connector deletion status is hung in Director Only .Can be additionally verified to make sure , no stale entry in controller through [INST]: Removing call home blacklist fot ZTP'ed connector in controller [/INST] : 1) Log in to controller ,(If its cluster setup login to VIP) 2) Get the list of blacklisted ips and select which ip would like to unblock curl -u : -k 'https://:8443/restconf/operational/adva-ensemble-controller:ensemble-controller/controllerconfiguration:callhome-config' --header 'Accept: application/json' --header 'Content-Type: application/json' NOTE:Replace the with the actual controller VIP and supply the controller rest username and password curl -u admin:admin -k 'https://10.64.107.192:8443/restconf/operational/adva-ensemble-controller:ensemble-controller/controllerconfiguration:callhome-config' --header 'Accept: application/json' --header 'Content-Type: application/json' 3) Delete the management ip from the callhome-blacklist Note: Replace the with ip-address from above curl -u : -k 'https://:8443/restconf/operations/baseservice:delete-stale-entries' --header 'Content-Type: application/json' --header 'Accept: application/json' --data '{ "input": { "stale-entries": [ "/adva-ensemble-controller:ensemble-controller/controllerconfiguration:callhome-config/controllerconfiguration:callhome-blacklist[controllerconfiguration:ip-address='\'''\'']" ], "data-store": "operational" } }' curl -u admin:admin -k 'https://10.64.107.192:8443/restconf/operations/baseservice:delete-stale-entries' --header 'Content-Type: application/json' --header 'Accept: application/json' --data '{ "input": { "stale-entries": [ "/adva-ensemble-controller:ensemble-controller/controllerconfiguration:callhome-config/controllerconfiguration:callhome-blacklist[controllerconfiguration:ip-address='\''198.21.7.177'\'']" ], "data-store": "operational" } }' 4)Get the list of blacklisted ips and check whether requested management ip deleted or not curl -u : -k 'https://:8443/restconf/operational/adva-ensemble-controller:ensemble-controller/controllerconfiguration:callhome-config' --header 'Accept: application/json' --header 'Content-Type: application/json' curl -u admin:admin -k 'https://10.64.107.192:8443/restconf/operational/adva-ensemble-controller:ensemble-controller/controllerconfiguration:callhome-config' --header 'Accept: application/json' --header 'Content-Type: application/json' [INST]: Troubleshooting Service delete Failures in Orchestrator [/INST] : Steps To Confirm & Resolve the problem: Signatures from Orchestrator showtech: From the orchestrator showtech(on all 3 orch nodes) go to the following location: \ESO\var\log\ensemble\eso-core and you can grep for the following signatures to troubleshoot the Service deletion failure on the orchestrator. 1) grep -i "" orch* > .txt Greps all the events and logs related to problematic UID from ALL orchestrator in the above folder and store it into a file. 2) grep -i "A deviceWithCloud already exists with name" orch* > .txt To find the errors confirming that cloud was already exsisits,grep for signature 2 from the output collected as part of signature 1 3)Signature to confirm service delete has been manually attempted from orch UI. grep -i "NorthBoundAPI:stopService" > .txt Check for the above signature in the logs grepped from 1 4) Signature to confirm that user has deleted/reset connector from director UI grep -i "NorthBoundAPI:removePhysicalDeviceWithCloud" > .txt Check for the above signature in the logs grepped from 1 5) If the user has used PDC scripts log file related to PDC script can be found in below location: \ESO\etc\opt\ensemble 6)Signature to confirm a delete has been intiated (from director UI) while an ongoing delete(i.e manual service deletion from orch UI) is in progress: [INST] - Mano Backup file to remote server procedure [INST]: In Director GUI - Connector callhome Status showing as status green however connector not reachable via SSH [/INST] : 1. Get or Check for Primary Database IP using cluster status command installer@app1:~$ clusterStatus 2. Access the primarary db server over SSH - ssh installer@ 3. Execute the below command in primarary db server i)Login into Database sudo /opt/adva/nc/mongodb/bin/mongo admin -u root -p 4netMGMT ii)Switch to Database use nc iii) Get the over all Operational Status of the UID - NOTE: In the command - substitute the UID db.getCollection("me_me").find({ "ctr.uid" : "900102939545-900102939543"}, { "mgmtip" : 1.0, "ctr.ctropsts" : 1 , "ctr.uid" : 1 } ); O/p would like below , CIB Status, Overall Oper & would show success iv) Update the status - CIB Status, Overall Oper & as failure db.me_me.updateOne({ "ctr.uid" : "900102939545-900102939543" }, { $set: { "ctr.ctropsts.chmstts" : "failure" ,"ctr.ctropsts.cibstts" : "failure","ctr.ctropsts.ovalstts" : "failure"} }) v) Now Verify the Dir UI if Oper State of UID updated as Down Kindly note - If the connector is restored/poweredup, automatically the states must transition back to Oper UP . This is confirmed by Etisalat as it was observed on one of the connector [INST]: Procedure to collect performance metrics from MANO Components [/INST] : Steps For Collecting Performance metrics from MANO components Director Performance Metrics • Login to director as installer • Copy monitor_dir.sh to /home/installer path • Add cron job for period collection • Add below entries and save the file (crontab -e) [INST]: Connector cloud status/physical device/Service Status/VNF State in Error showing as Unknown in orchestrator UI Because of Time shift in Connector [/INST] : Enter the Openstack controller Sudo /ovn/bin/dockersh ovn_ctrl Obtain the UUID of the server which is in the error state ‘openstack server list’ Verify VNF state If VNF state is in Error State ,then do the following nova --insecure reset-state --active Do an ‘openstack server list’ to check the vnf status Then Execute the following command: openstack server reboot --hard UID Do an ‘openstack server list’ to check the vnf status again [INST]- Unable to Import NST's(Specific ones) from Orchestrator to Director [/INST] : 1. Get all NSTs for Orchestrator using this API: GET: {{protocol}}://{{orchestrator-ip}}:{{port}}/orchestrator/api/networkServiceTemplates/list 2. Locate the required NST name and note down the key value (an integer). Example output: { "key": 2, "displayName": "Cisco_service_Fort", "deviceRequired": false, "cloudRequired": true, "networkServiceTemplateAllowChange": "any" }, 3. Next, run this API with the key; In the above example, the key value is 2. GET {{protocol}}://{{orchestrator-ip}}:{{port}}/orchestrator/api/ networkServiceTemplatesInternal/get/{{networkServiceTemplateId}} USAGE: GET {{protocol}}://{{orchestrator-ip}}:{{port}}/orchestrator/api/ networkServiceTemplatesInternal/get/2 4. Copy and paste the entire JSON output of this API response body to the notepad and name it as Cisco_service_Fort.json (based on displayName from Step-2 output) to your PC. 5. From the Director GUI, log on to the tenant. Create an NST template with name Cisco_service_Fort and use the file Cisco_service_Fort.json that you created in Step-4. [INST]- [Keystone Corruption ]Connector cloud status/physical device/Service Status showing as unavailable/down/Unknown in orchestrator UI [/INST] : Attached script is applicable where default credentials are used. Key notes - Check if it is for non-default credentials * security user name. 'security' is default * security user pass: 'security' is the default * Keystone admin password: 'ensemble' is the default Script attached. Need verification if ESO_STACK_DOMAIN_ADMIN_PASS is configured in the day1 before attempting recovery - Transfer script to Connector 'scp reset_openstack_passwords.sh security@:/tmp' - Access system via ovnchallenge - cd to home > c - copy script to home > cp /tmp/reset_openstack_passwords.sh . - Run the script ./reset_openstack_passwords.sh Script will ask for: * security user name. 'security' is default * security user pass: 'security' is the default * Keystone admin password: 'ensemble' is the default [INST] Orchestrator Polling Issue in 19.x.x due to missing eso-instance mapping [/INST] : Orchestrator Polling Issue in 19.x.x MOP to be applied on Orchestrator_cluster to resolve polling issue which leads to service edit failure. Steps to be followed: 1. Existing ZTP completed - While Inserting VNF trough service edit instance polling would report as not still booting. even when VNF is oper in connector 1. Login to all the 3 orchestrator via ssh 2. Run cluster_Status.sh on all the Orch cluster nodes 3. Access database on the Orch cluster nodes 4.Check within DB 4.a sudo mysql 4.b select id, name, last_polled_at, eso_instance_id from cloud where eso_instance_id; o/p would list N num of clouds without eso-instance ID. 5. Run stop_monitor command on all the Orch cluster nodes to disable VM monitor 6. Run systemctl stop eso-core command on Secondary node2,node1,followed by Primary node(VIP) 7.Then start all the nodes in a sequential way with a gap of 5 minutes after each node startup(Order should be VIP node,Sec-node2,Sec-node1) with the below command systemctl start eso-core 8.Check within DB on all the Orch cluster nodes sudo mysql select id, name, last_polled_at, eso_instance_id from cloud where eso_instance_id is NULL; select id, name, last_polled_at, eso_instance_id from cloud where eso_instance_id is not NULL; 9.Rerun the cluster status output on all the Orch cluster nodes cluster_Status.sh 10.Enable VM monitoring by running the following command on all the Orch cluster nodes to enable VM monitor start_monitor 11.Access the Orch UI.Perform the Service Edit & check if Issue resolves. [INST] - List New Job definitions/Image install fix post uprade path(19.1.x to 20.1.1/20.2./21.1) [/INST] : ESMBD-37939 TAC - Failed to install image using Director Staging(Image Download And Install) Job ESMBD-37903 Jobs: Post upgrade, new JOB definitions added in 21.x go missing 1. We can see that all 3 server have latest job_action.conf means 21.x version 2. But in mongo DB collection(jact) attributes are in 19.x version Steps followed ------------------- 1. do clusterStatus and note down the "Database Primary" Ex: Database Primary : 10.64.107.173 2. stop all 3 servers using aos_stop. 3. do aos_start in "Database Primary" node first. (Important step) Ex: aos_start in 10.64.107.173 4. start next 2 nodes with 2 min interval for each. 5. once all server are up we can see the new job type are listed in UI(login tenant view ->Job Definitions -> Create Job Definitions -> Type) Verify 6. New day n job instance was created and executed which completed successfully. 7. Jobs definitions/instance created in 19.x and migrated to 20.x failed (reason: because change in job definitions between 19.x and 20.x) [INST] - Cloud addition fails due to thread holding up old tenant context [/INST] : RCA: Tenant context in a thread is not getting cleared in case of service validation errors. That results in the thread retaining the old tenant context, which creates problem when the same thread is allocated for another tenant for cloud addition. Thread used in below service Validation on 27th March & still holding up Tenant context - BESIX.com orchestrator.log.32:2022-03-27 09:42:53,741 ERROR [THD:asyncCoreMessageListenerContainer-3|WKF:AddPhysicalDeviceWithCloudService|TEN:BESIX.com|CLD:|SVC:|VNF:|DEV:|PDC:mss-900102711914-900102711890-uEdge] (ServiceOrderPreDatabaseValidation.java:953) - Invalid Network CIDR value (host part is non zero): 10.110.2.105/29 orchestrator.log.32:com.overturenetworks.osa.api.exception.OSAException: Invalid Network CIDR value (host part is non zero): 10.110.2.105/29 When Programmatically above thread gets assigned for new cloud additions (7th April,20th April).As still it is pointing to BESIX.com & fails to fetch new tenant IDs i.e ALSHAYA.com while adding new clouds. orchestrator.log:2022-04-20 07:37:08,962 WARN [THD:asyncCoreMessageListenerContainer-3|WKF:|TEN:|CLD:|SVC:|VNF:|DEV:|PDC:] (AsynchronousCoreOperationReceiverImpl.java:37) - Could not fetch tenant with ID 19 orchestrator.log.23:2022-04-07 08:24:58,409 WARN [THD:asyncCoreMessageListenerContainer-3|WKF:|TEN:|CLD:|SVC:|VNF:|DEV:|PDC:] (AsynchronousCoreOperationReceiverImpl.java:37) - Could not fetch tenant with ID 160 Hence New Cloud addition fails. To overcome this issue - Fix in 21.1.1 - Programmatically handled, Thread will clear tenant context upon exit. Workaround until 21.1.1 1. Avoid CIDR misconfiguration on new activations - Invalid Network CIDR value (host part is nonzero): 10.110.2.105/29. 2. Eso-core restart on current VIP node, would free up Problematic thread state. Plan a maintence window to do a simple failover. stop_monitor sudo systemctl stop eso-core sudo systemctl start eso-core start_monitor Non-service impacting. Halt new activations. Access the Orch UI. Continue with further ZTP activations [INST] - Controller Peer Communication broken due to notification [/INST] : ESMBD-31536: TAC - ESMBD-37221 - Etisalat Deployments Blocked. Director Communication With Controller Down. When should TAC/end User run this MOP: 2022-04-07T13:51:06.328Z | com.adva.ensemble.app.adaptor.api.service.ActiveMQNotificationService | ERROR | 0 | on Executor: ssl:///10.183.173.207:61617 | Error on closing consumer for 10.183.173.207 class javax.jms.JMSException:Cannot send, channel has already failed: tcp://10.183.173.207:61617 Peer communicatin to controller is broken in UI Confirm below Ping to controller SSH to controller Rest curls [INST] - Update Mysql - connection.maxIdleTime & wait_timeout to address rollback exception [/INST] : ESMBD-31536 - TAC - Rollback Exception Occurs on Running Service,invokes recreation of Service & transitions to failed State When to Apply 1.Orchestrator Service Log must have traces - Can occur on Create & Edit Service flows. 2. The factors contributing to the issues are multifold : * Large NST(service order json). * Cloud environmental factors like disk latency (i.e. read and write wait time), which is making it little bit indeterministic : 20% failure rate. Procedure on Orchestrator Standalone System 1. sudo systemctl stop eso-core 2. sudo systemctl stop mariadb 3. In /etc/ensemble/eso-core/config.properties update the line connection.maxIdleTime=1500 to connection.maxIdleTime=3600 4. In /etc/my.cnf.d/server.cnf change "wait_timeout=1500" to "wait_timeout=3600" in [mysqld] section 5. sudo systemctl start mariadb 6. sudo systemctl start eso-core 7. Ensure that all the process are up. [ensemble@ensemble_orchestrator ensemble]$ ./etc/opt/ensemble/monitor_eso.sh [INST] - Monitor MANO Performance [/INST] : ESMBD-31536: TAC - Rollback Exception Occurs on Running Service, invokes recreation of Service & transitions to failed State. 1. Download the Tar & unzip. 2. Run the respective scripts on Respective MANO components. 3. Collect the CSV files generated & provide back to Engg Team for further analysis Procedure : 1. Share the Monitor output files to Engg Team for further investigations When should TAC/end User run this MOP: 1. Process failures i.e mysql, lwp, aos 2. Slowness observed in the Orchestrator and Director : Service Edits, NST payload (large number of variables), 3. MANO Engineer recommends to run this MOP for further investigation. [INST]: SW Rollback from 19.1.1 to 18.x [/INST] : Tracks [INST] deliverable, procedural steps and updated script. Comments : #1 by Tatiana Dietze on 2020-07-29 11:13 We should go here (ADVADocs for 19.1.1 MOPS but could use SalesForce for now if urgent before ADVADocs is ready. A Customer Permissions feature is in the works in ADVADocs. #2 MOP connections and Plan by Richard Jenny on 2020-07-29 20:29 Sunil Punathil - Prof Services MOPS contact. Derek Marks, Phil Penland in Atl - MOPS (Forms) and MOP expert SalesForce Retention: Sheena O'neill, Jens Bickel ADVADocs: Andy Spencer Today, we use SalesForce but ADVA Docs capability to be ready mid August to use ADVADocs. We should go here (ADVADocs) for 19.1.1 MOPS but could use SalesForce for now if urgent before ADVADocs is ready. A Customer Permissions feature is in the works in ADVADocs. [INST] Dell has created a MOP designated for use on the VEP4600 to increase the perform... [/INST] : Dell has created a MOP designated for use on the VEP4600 to increase the performance of the server (called "Turbo Boost"). This MOP has been created with modifications being done within the Grub settings and the CPU power states. This MOP needs to be verified against a server running 18.3.2.91 on the VEP4600 - version 146 was the GI for the VEP4600 where version 256 included the R640. Currently, an 18.3.2.91 GI is being created which will incorporate the changes that are called for in this MOP. MOP is enclosed. Please let me know if there are any MOPs you might have. [INST] - Director User Login Failure triage [/INST] : This MOP is meant to triage only User login failures NOT ALL USER LOGIN failures Logs path 1. showTech log path : 1. Application Logs : /Director_20240209_200147_IP_172.19.128.77_Rel.21.1.1_27/Director_20240209_200147_IP_172.19.128.77_Rel.21.1.1_27/Director_20240209_200147_IP_172.19.128.77/APPLICATION/Dir_Application_LOG.tar/Dir_Application_LOG/var/opt/adva/log/aos/nc 2. Keycloak logs : /Director_20240209_200147_IP_172.19.128.77_Rel.21.1.1_27/Director_20240209_200147_IP_172.19.128.77_Rel.21.1.1_27/Director_20240209_200147_IP_172.19.128.77/APPLICATION/Dir_Application_LOG.tar/Dir_Application_LOG/var/opt/adva/log/aos/nc/third_party 2. Below all across available Logs to be checked, app1_app-tier-scala-container* app1_C* 3. Signature to start with grep -i "Username" app1_app-tier-scala-container* 4. Try to find the exceptions, any errors in application logs & server logs LOWES/ETISALAT - Triage Live issues : UNABLE TO DELETE/CREATE Tenants/Connector Preconfig on Director UI [/INST] : How to Triage - Check with customer : Is it happening for all tenants/UID : Is it specific for one tenant/UID - Try to use Respective Director MOPs to clear Delete Hung states as a first step - If issue is happening across all tenants/UIDs ,check clusterStatus on Director , Controller - If Director cluster is non-operational , use aos_stop/aos_start on non-operational Node only else repeat aos_stop/start parallelly on all nodes - If controller clusters repeats non-operational , use famous Callhome QUESTIO - If both controller & Director shows cluster as non-operational , then for sure there was a momentary impact on sec-gw or outside which impacted call home on controller. Sometime controller auto recovers without call home MOP needed. We have check sec-gw journal log login as root to sec-gw nodes - master & backup check if any switchover journalctl | grep Entering - Director would auto revive on its own [INST] - Fix Opt disk size in health check script [/INST] : * Login to all Director cluster nodes * Edit the file sudo vi /opt/adva/nc/aos-nc-smartwan-installer/bin/health_check.sh In the below block USED_OPT=$(df -h | grep opt | awk '{print $5}' | cut -d '%' -f1) if [ $USED_OPT -gt 60 ]; then echo "Warning: $USED_OPT% of disk is used in /opt partition, db data compaction might help free up disk space." * Change from 60 to 75 * save the file * Rerun same steps on all nodes * Rerun the cluster Status script : cluster status should report all as operational/ No director Restarts involved [INST]: How to determine out of memory issues on the Controller (HEAPDUMP) [/INST] : Once heapdump is collected from a standalone or cluster then Engineering must analyze the logs. Look for the out of memory signature, if not present, heapdump can be helpful. Instructions: jps -m Get the Controller PID associated with "Main" Use PID in below command jmap -dump:live,format=b,file=/tmp/.hprof [INST] - Cleanup Journal Logs in Etisalat Production Sec-GW only [/INST] : 1. Post reboots , no new machine id is generated hence no reservation to new folders. It will be always one journal log folder with one machine id surviving reboot. Fix is available in 3.4.6b1 version 2. Log rotation of old archive logs - TBD which version [INST]: CIB DOWN in DIR due to DAYN PUSH using DAY1 File POST ZTP [/INST] : ISSUE - CIB DOWN in DIR due to DAYN PUSH using DAY1 File POST ZTP IMPACT - CIB DOWN , compute container as down fault in Connector Signature to Identify this 1.Check Event in Director UI for connector : Commissioning complete status /var/ansible/dayn.yaml Commissioning Complete FAILED : Access failed: Invalid Argument 2.Oper status of EmbeddedCloud Ready as Down. 3.Check below trace in the file - \var\log\commissioning_util.log Commissioning complete status /var/ansible/dayn.yaml Commissioning Complete FAILED : Access failed: Invalid Argument: Invalid IpAddress [400] : Invalid Argument: Invalid IpAddress 4. Check below error in \var\log\keystone\keystone.log Authorization failed. The account is locked for user 5. Identify openstack Corruption : Edit the file /etc/ovn/openstack_vars.json and verify below Attributes are set to Variables "ESO_CINDER_AZ": "{{hostname}}", "ESO_HOST_IP": "{{mgmt-ip}}", "ESO_CONTROLLER_IP": "{{mgmt-ip}}", "ESO_ORCHESTRATOR_IP": "{{orchestrator-ip}}", [INST] - Mariadb Recovery in Connector release 20.2.1 [/INST] : - showTech corrupted - With Engg Assistance , found Mariadb corruption from /var/log/mariadb/mariadb.log InnoDB: We intentionally generate a memory trap. InnoDB: Submit a detailed bug report to http://bugs.mysql.com. InnoDB: If you get repeated assertion failures or crashes, even InnoDB: immediately after the mysqld startup, there may be InnoDB: corruption in the InnoDB tablespace. Please refer to InnoDB: http://dev.mysql.com/doc/refman/5.6/en/forcing-innodb-recovery.html InnoDB: about forcing recovery. - Applied below workaround : Force recovery mariadb dockersh ovn_ctrl systemctl stop mariadb edit /etc/my.cnf.d/mariadb_openstack.cnf' under [mysqld] add: innodb_force_recovery = 1 systemctl start mariadb & exit (back to host) systemctl restart ovn_openstack_ctrl_newton - Verified the vnf status & service reverted to running , able to access VNC console [INST] - SSH troubleshooting on the VSR [/INST] : Please add all the knowledge learned from training and field issues. [INST]: LTE FW upgrade [/INST] : #1 by Tatiana Dietze on 2022-08-23 21:21 We should not be responsible for doing FW upgrade. It can be done by the customer and carrier via OTA. [INST] - Debug the issue when BGP is lost and IKE is up [/INST] : How to debug on 6wind a)Login to master and check the following 1. Check the status by executing the command: "show state vrf main vrrp" and "show state vrf mano vrrp" -- the status should be master in both. Kindly Note - It should not have backup or fault 2. check ike connectivity: "show ike ike-sa" -- all the tunnels should be in established status 3. capture the ike count: "show ike ike-sa-count" 4. execute command: "show bgp neighbors" --- This shows zero. as BGP neighbors are lost. b)Login to backup and check the following: 1. Check the status by executing the command: "show state vrf main vrrp" and "show state vrf mano vrrp" -- the status should be backup or fault in both. 2. check ike connectivity: "show ike ike-sa" -- all the tunnels should be in passive status 3. capture the ike count: "show ike ike-sa-count" Workaround to recover the BGP: Toggle ike in the master where ike connection is established. edit running vrf main ike enabled false commit vrf main ike enabled true commit This will bring up the BGP and ike in master. [INST] to be executed on Controller 20.1, post MANO upgrade [/INST] : This MOP should be executed only once on a Controller terminal, once MANO has been upgraded to 20.1. * Upgrade the MANO from 19.1.1 to 20.1.1 * Make sure don't perform any operations(especially Reset, Delete) after upgraded to 20.1. First Apply below steps before doing any operations. * Please find the attached jar file which creates the recovery-readiness model in 20.1 controller datastore, when the data is migrated from 19.1 to 20.1 * Ssh to Controller terminal ssh ensemble@ * example : unzip EnableDisableRecoveryFeatureOkHttp.zip * It should be run on the controller instance after 20.1 controller is up * java -jar EnableDisableRecoveryFeatureOkHttp.jar admin * example : java -jar EnableDisableRecoveryFeatureOkHttp.jar admin admin * Note: if Controller is cluster Execute the Jar in VIP Node once controller Cluster is up and running. if the REST 'admin' user password has been changed, please use the new password in step number 6, last parameter. 7. While Creating the Connector, Explicitly set the enable Recovery(enrec) and Create Service(crsv) flags while using API 8. In case Tenant was created without enable Recovery(denrec) and Create Service(dcrsv) flags , Update the tenant Global settings and provide value for both "denrec" and "dcrsv" as patch request for tenant in the API flow [INST] - Clean up stale data upon deleting only connector preconfig post 20.1.1 Upgrade [/INST] : Polarian ESMBD-35015 This MOP is applicable only when below conditions are met 1. Connector preconfigs are created in 19.1.1 with NO ZTP done 2. UPgarde MANO to 20.1.1 3. Delete is attempted from director on connector preconfig created in 19.1.1 To remove such above stale entries 1. Download the staledelete jar to controller system . Can be executed from any path 2. Usage: java -jar ConnectorStaleEntriesPurger_NoConfig.jar