Datasets:

kristmh

/

mongoDB_testset_highest_high

like 0

the pidfile appears to be written even in the case where the database cant be started due to lockfile being lockedeg mongod quiet dbpathbackendscriptsdevdb logappend logpathdblog pidfilepathdbpid forkforked process cat mongod quiet dbpathbackendscriptsdevdb logappend logpathdblog pidfilepathdbpid forkforked process cat ps aux grep mongobenno s mongodi think it would be better if in the case where the lockfile cant be obtained the pid file isnt updated ie the pid file should be protected by the lockfile

the error returned is illegal instruction should fix this by troubleshooting the parameter we pass to g visible here and we should prevent this by testing on a machine after building on a machine

the version of openssl bundled with python the last version binary windows installers exist for appears to be there is no way to tell from the ssl module itself in python but a note in the changelog mentions that version attempting to test with requests fails with alert protocol version noformat c import requests printrequestsget verifyfalsejson snimissingwarning an https request has been made but the sni subject name indication extension to tls is not available on this platform this may cause the server to present an incorrect tls certificate which can cause validation failures you can upgrade to a newer version of python to solve this for more information see snimissingwarning insecureplatformwarning a true sslcontext object is not available this prevents from configuring ssl appropriately and may cause certain ssl connections to fail you can upgrade to a newer version of python to solve this for more information see insecureplatformwarning traceback most recent call last file line in file line in get return requestget url paramsparams kwargs file line in request return sessionrequestmethodmethod urlurl kwargs file line in request resp selfsendprep sendkwargs file line in send r adaptersendrequest kwargs file line in send raise sslerrore requestrequest requestsexceptionssslerror httpsconnectionpoolhost max retries exceeded with url acheck caused by alert protocol version noformat the reason the failures have started with mongodb is that version disabled tls by default openssl didnt add support for tls until version we have two options dont test the combination of python tls windows and mongodb explicitly reenable tls using the command line option on mongodb

this page could use a toc ideally the fields in the example document could link to more info about that field but that may not be possiblesee

modify the health check section to mention the debug port will only listen on the loopback interface and it is not configurable

in debugging an authrelated failure today i came across the following message in mongod log i access sasl authentication failed for dev on admin from client authenticationfailed unable to perform scram authentication for a user with missing or invalid scram credentials this message conflates two nonoverlapping failure modes the credentials were missing and thus scram authentication was not attempted credentials were supplied authentication was attempted credentials were found to be invalid each of these failure modes should have its own separate log message i used a server for the test but master appears to have the same message string in it

by default mongod does not listen on however localhost may resolve to an address on some machines which could lead to a connection failure as discussed in changing localhost to should avoid this note this is still a documentation issue as users will need to be careful about using hostnames that resolve to addresses unless is enabled in the mongodb configuration

were experiencing regular index corruption on a table with a partialindexfilter though it is not always that index getting corrupted the only relevant thing i could find near the time when the corruption is occurring is log rotation via nothing else in the log file seemed suspect unfortunately our data is very sensitive and cannot be shared i can tell you were using rubyrails and the mongoidparanoia gem and set up a unique partialfilterexpression with code v unique true key name name uniqueactivenames ns apiproductionappinstances partialfilterexpression deletedat null code the corruption appears to have only been exposed with this added but may have happened before and gone unnoticed we found it when being unable to find documents with this index using code deletedat null name name code

in various situations eg testing against atlas clusters or with clusters started with mlaunch a mongodb cluster is started without support for configurefailpoint enabled it would be more convenient if tests that rely on configurefailpoint were skipped rather than failed

description gcb gcp gcp grbi cw r cc backport to wq ggf scope of changes impact to other docs mvp work and date resources scope or design docs invision etc

when generating keystring keys we use the keystringpooledbuilder that places multiple temporary keystring on the same larger memory block to avoid multiple small allocations each keystring holds a reference to the underlying memory that will not be freed until all keystring that is using it has been freed normally this is fine as these keystring are passed to the sorter that spills do disk when the memory consumption reach a certain threshold as this process clears all temporary keystring instances all memory blocks should be freed however when there are multiple indexes being built at the same time these pooled memory blocks can be shared between keystring instances belonging to different indexes if some indexes generate large keys and need to be flushed to disk often the actual memory will not be freed if there is still an index building that hasnt needed to spill to disk

due to the findings in a node entering quiesce mode in a mixed replica set could delay a node finding a valid sync source some options to prevent this are to backport the changes in to or only enable quiesce mode when the featurecompatibilityversion is

hi this page mentions that the root resource is at get however that is the cloud manager public api the atlas api seems rather to be at get the examples further down the page are correct and use the endpoint

millioncollectiontest failed on large scale testshost wiredtiger developcommit diff fix macos build jun utcevergreen subscription evergreen event task logs

createindexglejs started intermittently failing since yesterday afternoon the failure symptom is that a getlasterror command returns a could not target full range of testuser metadata not found i am unable to reproduce this failure on my desktop a list follows of recent failures linux ssl ubuntu ssl amazon linux ssl rhel excerpt from first failure in above list noformat d user assertion master and slaveokfalse d command assertion while executing command listdatabases on database admin with arguments listdatabases and metadata configsvr optime ts timestamp t not master and slaveokfalse i command command admincmd command listdatabases listdatabases exception not master and slaveokfalse locks protocolopcommand i sharding distributed lock with ts unlocked w sharding could not initialize targeter for index write op in collection testuser i sharding distributed lock test acquired ts d user assertion master and slaveokfalse d command assertion while executing command listdatabases on database admin with arguments listdatabases and metadata configsvr optime ts timestamp t not master and slaveokfalse i command command admincmd command listdatabases listdatabases exception not master and slaveokfalse locks protocolopcommand i sharding distributed lock with ts unlocked w sharding could not refresh targeter caused by totalsize field not found in listdatabases w sharding failed to get database config for testsystemindexes while checking for autosplit namespacenotfound database test not found assert write or write concern failed err could not target full range of testuser metadata not found code n shards shardrawgle ok e query error write or write concern failed err could not target full range of testuser metadata not found code n shards shardrawgle ok failed to load jstestsglecreateindexglejs noformat

opsmgr only and by group owner role user

a few months ago i tested communicating with the mac hosts using grpc instead of ssh because ssh is fragile and highly dependent on the ssh command producing wellformed output im pretty sure i also filed a ticket to allow the app servers to reach out to jasper on the mac hosts over grpc but it didnt work even after the ticket was closed since the annoyance around relying on ssh output has arisen again recently its worth trying again and diagnosing why exactly grpc isnt working

the correct authentication requirements arenoformatdbadduser user username pwd password roles otherdbroles local admin noformat

ive been on forums talking with several people for the last few days and no one has been able to give me an answer is there anyway to reconfigure a replica set in the event the primary as well as a majority of the servers are down for example if data center goes down a procedure can be run to reconfigure and set a new primary with the remaining serversi tried the eval method on the database object and i get thiscommand eval failed not master response note from execcommand ok errmsg not master i also tried thisvar reconfigcommand new commanddocument replsetreconfig new bsondocumentconfiguration newconfigforce truevar response databaseruncommandreconfigcommand and runcommand throws the following mongocommandexceptioncommand replsetreconfig failed replsetreconfig command must be sent to the current replica set primary response ok errmsg replsetreconfig command must be sent to the current replica set primary from what i can tell in the documentation force true should allow reconfig against a nonprimary databaseany suggestions

objectids equals method is implemented invalidly equals methods need to be symmetrical which means if aequalsb is true bequalsa needs to be true as well this is not the case as the implementation massages string values into objectid instances and thus surprisingly the following code succeedscodejavaobjectid left new objectidstring right lefttostringassertthatleft isright succeedscodethis breaks once you test for symmetrycodejavaassertthatright isleft failscode

dbcheck currently takes a modes collectionlevel lock this can likely be changed to improve performance

depends on “refactor inputs to host allocation” requirements respect the distro scheduler flag per distro toggle respect the respect the top level host allocator flag schedulergethostallocatorctx schedulerhostallocatordata int err bulk insert intents probably can be disabled to start as this depends on the scheduler part

code mon jul cmd drop jul got signal segmentation faultmon jul invalid access at address from thread mon jul backtrace code

headerline is ignored if you specify typejson or leave it as default but it is meaningless unless you are using csv or tsv so it should be an error

for tutorials that require starting shard replica set members in standalone mode ensure that we specify removing shardsvr and readding it when adding the member back to the shard replica set

the path for the gotools directory has changed quite a lot between and homebrew and probably other package systems are failing due to this the following path has changed to it appears to be a copy and paste error perhaps compare with

hii was trying to get the rename code working and i think i stumbled across a possible issue with the example create person class with name and age serialized an instance of person to my code returns collection of person and displays to changed person class to name and oldage serialized an instance of person to my when attempting to display collection to console it exceptscodepublic class person isupportinitialize public objectid id get set public string name get set public int oldage get set public idictionary extraelements get set public void begininit public void endinit object agevalue if extraelementstrygetvaluealive out agevalue return var age intagevalue extraelementsremoveage oldage age problem when deserializing the second record extraelements is nullmy resolution the only way that i could make that code work was to do either of the add a constructor to new up extraelementspublic person extraelements new check if extraelements is nullif extraelements null extraelementstrygetvalueage out agevaluereporter richard oneilemail

in overview section there is a typo in replication introduction link phrase which is written as repication introduction please fix it

once is complete a mongos process will quickly exit with an error status if you point it at config servers with featurecompatibilityversion we should write a test in the multiversion suite to verify this behavior the test should perform the following steps start an all sharded cluster verify that the feature compatibility version is assert that a new mongos fails to start up when joining the cluster set the feature compatibility version to assert that a mongos can now start up against the cluster and can perform reads and writes to the shards

right now it is too vague on how tags selection worksbqtaken from the speca member matches a tag set if its tags match all the tags in the setfor example a member tagged dc ny rack size large matches thetag set dc ny rack a members extra tags dont affect whether its a matcha read preference optionally includes a list of tag sets a driver or mongos searchesthrough the list from first tag set to last looking for a tag set that matchesone or more members which also match the mode

the following code does not compile due to ambiguous bsonobjbuilderappend overloadcode orz bsonobj o bsontest orzgcc compile error error call of overloaded appendconst char long unsigned int is note candidates are void mongoappendconst char mongobsonobj note void mongoappendconst stdstring mongobsonobj note void mongoappendconst char note void mongoappendconst char note void mongoappendconst stdstring note void mongoappendconst char unsigned note void mongoappendconst char long long note void mongoappendconst stdstring long long note mongobsonobjbuilder mongoappendconst char note void mongoappendconst char note mongobsonobjbuilder mongoappendconst char const char note void mongoappendconst char stdstring apparently the problem is that bsonobjbuilderappend does not overload specifically for unsigned long long type so for its possible to match all types leading to ambiguous overload this bug is in all release including the trunk version

the cursor implementation in the mongocdriver makes this assumption if the server returns a nonzero cursor id then the cursor isnt finished because there are more documents to iterate through right now sharded clusters return a nonzero cursor id even when the document limit has been reached so our cursor throws an error when it gets conflicting information the userdefined limit has been reached but the cursor id isnt zero here is the command that fails all the way back to mongocdriver version code mongoccollectionfindwithopts collection bsonnew filter bconnewlimit batchsize other options null code here is the assertion that fails when the command is ran code remaining limit cursorcount bsonassert remaining code i attached an exampleclientc file that can be ran to reproduce the bug

this issue has been posted on stackoverflow the problem is that with net i am not able to use custom bsonconstructor with more than parameters whereas if i use net the issue is not here i search a long time a workaround for this issue and i dont have any idea of what causes this

the goal of this ticket is to remove any references to commands which stopped being used in but still need to be present for compatibility from the js tests when becomes the last lts version

we have a mongodb replica set consisting of servers arbiter all running mongodb with the wired tiger storage engine i updated the secondary to it appeared to start syncing from the primary the state transitioned to secondary and the logs said it was syncing from the primary but crashed after about seconds with this output in the logs noformat e storage wiredtiger sessioncommittransaction memory allocation cannot allocate memory i invariant failure scommittransactions null resulted in status unknownerror cannot allocate memory at srcmongodbstoragewiredtigerwiredtigerrecoveryunitcpp i control begin backtrace mongodbversion gitversion uname sysname linux release version smp thu jun utc machine somap elftype b buildid b elfty pe buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid mongod end backtrace i aborting after invariant failure noformat i restarted mongodb and the same thing happened i downgraded to mongodb and the secondary was able to rejoin the replica set and catch up as normal the secondary wasnt under any read load at the time but the primary was under a heavy write load

we have it for optionscount but curiously not optionsfind this is pretty important to users who have lots of indicesnontrivial queries and want to be sure the right plan is selected

description hi there ive just detected that create one restore job for one cluster example request page is currently listing an example which will never work due to a typo current example for http restore code curl user usernameapikey digest header accept applicationjson header contenttype applicationjson include request post data delivery expirationhours maxdownloads methodname http snapshotid snapshotid code should be missing comma sign after maxdownloads code curl user usernameapikey digest header accept applicationjson header contenttype applicationjson include request post data delivery expirationhours maxdownloads methodname http snapshotid snapshotid code could you please have a look and fix thanks a lot in advance alexey scope of changes impact to other docs mvp work and date resources scope or design docs invision etc

ive a mongo cluster version with config nodes data nodes and router the router frequently goes down almost times a day with the below error noformat i access authenticate db azuga authenticate user azugaapps nonce xxx key xxx i sharding cluster pinged successfully at thu nov by distributed lock pinger sleeping for i network connection accepted from connections now open i access authenticate db azuga authenticate user azugaapps nonce xxx key xxx f got signal aborted begin backtrace mongodbversion gitversion uname sysname linux release version smp thu mar utc machi ne somap elftype b buildid b elftype buildi d b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elftype buildid b path elft ype buildid b path elftype buildid b path elftype buildid b path elftype buildid mongos mongos mongos end backtrace noformat below is the detail of dbstats on the major db where we have of data noformat mongos dbstats raw db azuga collections objects avgobjsize datasize storagesize numextents indexes indexsize filesize nssizemb extentfreelist num totalsize datafileversion major minor ok db azuga collections objects avgobjsize datasize storagesize numextents indexes indexsize filesize nssizemb extentfreelist num totalsize datafileversion major minor ok objects avgobjsize datasize storagesize numextents indexes indexsize filesize extentfreelist num totalsize ok noformat

for queries that include a sort operation without an index the server must load all the documents in memory to perform the sort and will return all documents in the first batch sounds wrong

the bulk write api added in supports sending inserts updates and deletes in bulk to the server if the number of writes requests exceeds or the total size of the write requests exceed the driver will split the write requests into batches of andor batches whose size does not exceed ordered bulk writes the driver stops sending more batches to the server if there were any errors reported in the current batchfor unordered bulk writes the driver should keep sending the rest of the batches to the server even if errors were reported in the current batch there is a bug which is preventing that from happening instead the driver stops sending more batches and immediately reports the first batch of errors to the application ignoring any subsequent batchesan application will be effected by this bug in the following circumstances it is using the method dbcollectioninitializeunorderedbulkwriteoperation the number of writes added to the operation exceeds or of total data the possibility of an error most likely a duplicate key error existsworkaroundlimit the number of writes in a single bulk write operation to less than and less than mb

given the example of a set of documents like thiscode javascript status a type c status a type d status b type c status b type d codecurrently if you wanted to find all documents that did not contain the combination of status a and type c you would have to resort to javascript evaluationcode javascript dbcollectionfindfunction thisstatus a thistype c codethe aggregation pipeline allows such casting to a projection of a fieldcode javascript dbcollectionaggregate project status type unmatched not and eq eq match unmatched false codejavascriptbut as there is not a native form of this for a query you could not filter for a match stagethe seemingly natural form would be the similarcode javascript dbcollectionfind not status a type c codewhere the existing not operator applied its left side evaluation to the query match negating the whole expression of the query on the right sidenot sure if there is an efficient way to apply this to index values or whether the result would be the same collection scan for all documents not matching the exclusionthe general problem presently being that not is not allowed as a top level operator here at the moment and cannot reverse the entire query condition as it expects a left side field value and the javascript form for a where clause is not allowed within an aggregation expression noformaterror err cant canonicalize query badvalue unknown top level operator not code

this may be related to the recent change to allow private project variables without this fix it may not be possible to create projects with the current ui

abstract similar to setsudo as follows in pre step command expansionsupdate code key activatevirtualenv value check if virtualenv is set up if then if then venvscriptsactivate else venvbinactivate fi fi code invoke as activatevirtualenv

lets just remove all the breakage since its from days

deviceresource busy error was raised during long running schema test in for job wiredtigertestunitlong on it does not look like fallout from which changes csuite test though theres time proximity and both are around schema testing noformat error subunitremotedtestcase stringexception traceback most recent call last file line in testmanytables selfsessiondropuri wiredtigererror device or resource busy noformat

in our documentation we say that the find actiontype grants you permission to run listcollections say we create a custom role that grants the find actiontype on some database code dbruncommand createrole findrole privileges roles code a user with the findrole role will not be able to run listcollections this is subtle its because specifying empty string as the collection for a resource excludes system collections and we require permissions on systemnamespaces to run listcollections while nothing in our documentation is strictly wrong it is certainly misleading

sorry if this is the wrong place to put this this is a major problem because even an install script that is specifically attempting to install will appear to succeed even though at present will actually be installedhere is a transcription copied from my terminal to demonstrate the issuecode sudo yum install plugins fastestmirror refreshpackagekit securityloading mirror speeds from cached hostfile base mirrorsadamsnet epel linuxmirrorsesnet extras centosmirrorjchostnet updates mirrorthelinuxfixcomsetting up install processpackage is obsoleted by mongodborgserver trying to install insteadresolving dependencies running transaction check package will be obsoleted package will be obsoleting finished dependency resolutiondependencies resolved package arch version repository sizeinstalling mongodborgserver m replacing summaryinstall packagestotal download size mis this ok cexiting on user command sudo yum install plugins fastestmirror refreshpackagekit securityloading mirror speeds from cached hostfile base mirrorsadamsnet epel linuxmirrorsesnet extras centosmirrorjchostnet updates mirrorthelinuxfixcomsetting up install processno package availableerror nothing to do sudo yum install plugins fastestmirror refreshpackagekit securityloading mirror speeds from cached hostfile base mirrorsadamsnet epel linuxmirrorsesnet extras centosmirrorjchostnet updates mirrorthelinuxfixcomsetting up install processresolving dependencies running transaction check package will be obsoleted package will be obsoleting finished dependency resolutiondependencies resolved package arch version repository sizeinstalling mongodborgserver m replacing summaryinstall packagestotal download size mis this ok cexiting on user commandcode

we will release a single version of the tools for each platform which will support both the community and the enterprise edition of mongodb as such we should have a single buildvariant per supported platform that builds and tests this single release

hello this problem can be easily solved in a naive and thus probably very inaccurate way by wrapping the pymongo findupdateremoveinsert methods with time measuring function but the driver itself would be the ideal place to put some kind of instrumentation thanks for considering

if you start a scopedthread in a test and it throws an exception that exception is swallowed and does not error the test even if the main test thread calls join and returndata on the scopedthread this can lead to subtle problems with tests and issues where tests are broken and no longer testing what they are supposed to but no one notices

this issue is in reference to libmongodbscopejsthere is a constructor function cursor which defines methods every time the constructor is called these methods should be moved to the prototype and the scopeoptions and cursor variables should be moved to instance variablesa lot of the defined methods just callthrough to the cursor object which makes me believe cursor should inherit from it looks like somebody else had this idea as well as i can see requireutilinherits at the top of the file but it is never used anywherefixing the inheritance chain would also make it possible to promisify the cursor for example promisifying the cursor constructor with bluebird right now doesnt do anything because cursors returned by the api use this internal cursor function which is not exposed anywhereit also appears that on line fields is leaked to the global scopei am willing to do the work to fix these issues if nobody else is working on it i would like to fix it quicklyjust for reference other people are also running in to issues

doing a backup by doing an fsynclock and then copying the dbpath files is not a supported approach for mongod instances running wiredtiger we need to document this in release notes when fsynclock is mentioned possibly elsewhere

tried to shutdown a secondary that has considerable repl lag calls to shutdownserver seem to hangbacktrace attached

download button for ubuntu doesnt work

update etcperfyml on master and to use latest mongoperf release

currently we use the primary read preference for all aggregations due to the complexity of checking an interface for out once we have an insertionordered map type implemented we should check the pipeline for out and select a server and read preference accordingly

html drivers ticket description script target if you can read this text the script has failed get functiondata var description datafinddescriptionval langscripttargethtmldescription html

the changes from made it so evergreen tasks which run server tests now depend on archivedisttest binaries rather than archivedisttestdebug debug symbols the evergreen fetch command run while spawning an evergreen host only downloads the artifacts from transitively depended on tasks this means the debug symbols tarball is no longer automatically available to engineers and the setupspawnhostcoredump script when spawning a host

when trying to parseint on a bson numberlong the result will be nan instead of a number

fsm test fails in a replicated environment with this failure the fsm test interleaves updates deletes inserts and querying with multiple threads the query used is specifically on id to use the idhack query phase f invalid access at address f got signal segmentation fault begin backtrace backtraceprocessinfo mongodbversion gitversion uname sysname linux release version smp tue mar edt machine somap mongod mongod end backtrace noformat

create an ecs client and a wrapping manager to wrap the ecs calls the ecs client should interface with the ecs sdk and return raw ecs results maybe with some logic to retry and handle errors specially the manager should just be a wrapper to help make calls with the ecs client we also need a client for secrets manager

when i request a project that i am not an admin of using the api i receive a even though i provided credentials evergreen get failed as a user of evergreen api i expect the api to use correct http status codes for its responses so that i have an accurate understanding of the error condition since i provided credentials and the server is denying access i expect to receive http code rather than

on solaris we run into a blocker when using bsonextits exactly the same problem as described here it would be great to have this modification included into

i have an issue that is rather difficult to debug and was wondering if anyone sees anything wrong with my configuration i posted an issue in the mongoose github repo but have not had any luck there was hoping someone could help me figure this out error no primary server available nodejs version and mongodb version with mongoose this seems to happen randomly and will open many connection until i finally restart the node process the cluster is healthy at all times during this error this error happens hundreds of times per hour there does not seem to be any consistency as to when the error will begin for example it occurs when the cluster is operating normally and no changes to the primary have been made this is what the db stats look like as you can see the number of connections will steadily increase if i kill the node process and start a new one everything is fine see attached image config connect mongooseconnectconfigmongoconnectionstring server socketoptions sockettimeoutms keepalive replset socketoptions sockettimeoutms keepalive connection string stack trace pickserver command command executecommand command findandmodify findandmodify anonymous function findandmodify findandmodify findoneandupdate none none while connections are on the rise the status of the cluster is healthy set date mystate members id name health state statestr secondary uptime optime optimedate lastheartbeat lastheartbeatrecv pingms syncingto configversion id name health state statestr primary uptime optime optimedate electiontime electiondate configversion self true id name health state statestr secondary uptime optime optimedate lastheartbeat lastheartbeatrecv pingms syncingto configversion ok i have been outputting mongoose connection events and it appears to try and reconnect a lot ​ mongoose event fullsetup ​ mongoose event reconnected ​ mongoose event open ​ mongoose event connected ​ mongoose event reconnected ​ mongoose event reconnected ​ mongoose event reconnected ​ mongoose event reconnected ​ mongoose event reconnected ​ mongoose event reconnected ​ mongoose event fullsetup ​ mongoose event connected ​ mongoose event open ​ err name mongoerror message no primary server available

testrepotracker failed on ubuntu evergreencommit diff evergreencievergreen pull request by kimchelly avoid panic if dialing cedar errors may utcevergreen subscription evergreen event task logs

see for detailed description and log files quote my use case involves consumers pulling data from two queues and making inserts into two other large collections as adding data to a capped collection where i log the jobs im using ubuntu and have attached my logs here any help would be truly appreciated if i cant figure this out sometime soon im afraid ill be forced to return to mongo i never encountered the entire database crashing and having to be manually restarted every couple hours before quote noformat i control begin backtrace backtraceprocessinfo mongodbversion gitversion uname sysname linux release version smp fri aug utc machine somap end backtrace i aborting after invariant failure noformat

test standalone mongod test with following benchrun scriptcodevar run function var d dbgetsiblingdbtest res benchrun ops ns testfoo op update query id update inc a w writecmd true seconds totals true writecmd true w parallel return resvar s dbserverstatusvar v dbserverbuildinfoprintversion vversionprintgit version vgitversionif sstorageengine undefined printstorageengine dbserverstatusstorageenginenameprintupdate rps results the throughput from both wt and is slower than that of test wt contenedupdate update with thread update with thread observation wt update is slow this probably related how update is done with wt with single thread update it is slower than and with thread update is slower when there is contention and in par with single thread update

this should be in the release notes or linked from there at least

codejava var find collectionfindx xarrayfieldanypredicate var a findproject buildersprojection includex xarrayfield excludex xid firstordefault ok arrayfield var b findprojectx xarrayfield firstordefault mongointernalexception unexpected fieldcode both a and b should be the same but b throws an exception if firstordefault is replaced with tostring both queries are exactly the same there is a bug in the projection object deserializer

the fields of keepalivevalues should be uint bytes instead of ulong bytes see the ulong types referenced here are in the c language where ulong is bytes the reported side effect of this bug is that hundreds of keep alive packets are sent per second

typically when specifying a field in a query with a mix of operators or exact matches you will get an error indicating that the operator is unknown this error is correct in that it would be ambiguous to accept the mixed format when considering how exact matches work for values in this syntax expected error code test dbtestfindsubfield lt a error error ok errmsg unknown operator a code codename badvalue code however specifying a nonoperator key first allows one to bypass this error to get behavior that seems like an exact match where the field is including in the keys to match failing case code test dbtestfindsubfield a lt fetched records in code

as we have transitioned to using the uuidcatalog as authoritative collection catalog it makes sense to update its name to reflect the new role

this pages discusses a geo query for but that should then be a negative number

how does one figure out that our latest mms release is unless you know that you cant download the latest mms version

this can help to produce inner join by intersect two or more mapreduce outputs

when compiling mongoc with cmake then both the normal shared library and the static cmake config is required to be available this is despite the fact the static config is only installed when explicitly configured to do so if libbson is built with cmake then it always builds and installs both static and shared libs so i think this will be a problem only in the strange event that libbson is built with autotools which doesnt build and install the static libbson by default and then libmongoc is built with cmake plan done add enablestatic option to libbsons cmakeliststxt default on done add enablestaticonoffauto option to libmongocs cmakeliststxt default auto done if enablestatic is on require static libbson if auto and static libbson not found set off if off dont search for static libbson done update libmongocs linksampleprogramsh test to pass denablestaticoff when compiling libbson and libmongoc if the env var linkstatic is not set

does not explain what underlying mongodb roles the custom atlasadmin role grants

the mongodb manualpdf i read the http links in that book are all wrongcause the manualreference has been changed to manual in itso i wish you could fix itthx

mongodb fails to start when provided an certificate whose subject name is empty windows provides us with the number of subject name rdns as an unsigned integer from which we subtract the arithmetic on value line is going to be on unsigned types unsigned arithmetic has defined overunderflow semantics will evaluate to a large number which can be correctly assigned to i

i installed pymongorootbeniovarlog easyinstall pymongosearching for pymongoreading best match pymongo processing q bdistegg distdir flag not set analyzing archive contentsadding pymongo to easyinstallpth fileinstalled dependencies for pymongofinished processing dependencies for pymongoafter that i run pythonpython next from gridfs import gridfsand got an errortraceback most recent call last file line in file line in file line in importerror cannot import name nextitem

the feature tracker is not part of the durablecatalog abstract interface when interfaces was moved from durablecatalog to collection in it resulted in prepareforindexbuild having to call multiple functions on the durablecatalog because some work needs to be adjusted depending on the featuretracker featuretracker should be deleted altogether

there are several checks for whether twophase index builds are enabled however we never actually run the test with singlephase index builds enabled we should modify the test to expand test coverage to guarantee our timestamping logic is correct

there are two different functions used to allocate ports in jstests one of them actually keeps track of what ports have been used and allocates the next one but the other does not and just allocates a block that you request without any checks the replsettest helper uses the one that does not do any checks which means that two replsettest helpers cannot be used simultaneously without allocating and setting ports manually the other helpers dont have this problem except for shardingtest which uses a replsettest object when the shards are replica sets

symptom noformat dbtimelinecacheupdateid in push c each slice update q id in u push c each slice multi false upsert false does not contain id or shard key for pattern id hashed writeresult nmatched nupserted nmodified writeerror code errmsg update q id in u push c each slice multi false upsert false does not contain id or shard key for pattern id hashed noformat the same update with works fine arguably could be interpreted during upsert as equality

i can successfully compile and link when using denablestaticon when running the executable on a different system it still tries to load a dynamic libbson error while loading shared libraries cannot open shared object file no such file or directory

we had a failure of testschemaabort in this job it ran on tinderbox and threads are stuck waiting for io to complete changes were made to timestampabort that got rid of the stableperiod and simplified the timestamp handling in similar changes to schemaabort should be made

description hi there is specification missed in the below document regarding collstats output collstatscusorwhat is the meaning of all the terms including in collstatscursor output like reset calls create calls insert calls remove calls please provide the meaning of all these scope of changes impact to other docs mvp work and date resources scope or design docs invision etc

as an evergreen engineer when a user visits the evergreen website then id like to know how long it takes that user to loadrender the entire page

we want the currentop entry for both active and inactive transactions to contain an entry with a transactionparameters sub document that contains the timeactivemicros and timeinactivemicros of the transaction we should be able to retrieve these real time values by using the singletransactionstatstimeactive and singletransactionstatsstarttime fields stored on the session

right now if you try to build with usesystempcre the include path is still expanded with the include path to reference the vendored pcre path include paths to normally vendored libraries should not be added to the build line when the system library is requestedthis only affects

first question complicates the answer by mixing locking behavior and granularity of locks this should be split out and lock def should link to wikipedia page with better description in wiki docsyielding happens for more than disk access which was added in and expanded in multiple document writes yield as well as readers periodically this can allow reads during long multidoc write operationsand there are more issues here so it is best if this put back on the list for more review

we require to have driver version in our projects but while installing it there are errors coming in envc file as below please let me know how can we fix this cc o srcenvos c pedantic wall ggdb dposixsource dmongohavestdint fpic dmongodllbuild srcenvc srcenvc in function ‘mongoenvsocketconnect’ error storage size of ‘aihints’ isn’t known struct addrinfo aihints warning implicit declaration of function ‘getaddrinfo’ status getaddrinfo host portstr aihints ailist warning implicit declaration of function ‘gaistrerror’ bsonerrprintf getaddrinfo failed s gaistrerror status error dereferencing pointer to incomplete type ‘struct addrinfo’ for aiptr ailist aiptr null aiptr aiptrainext warning implicit declaration of function ‘freeaddrinfo’ freeaddrinfo ailist warning unused variable ‘aihints’ struct addrinfo aihints recipe for target srcenvos failed make error

under some load concurrent threads the driver throws the following exception message unable to connect to a member of the replica set matching the read preference primarystacktrace at mongodbdriverinternalmultipleinstancemongoserverproxythrowconnectionexceptionreadpreference readpreference at mongodbdriverinternalmultipleinstancemongoserverproxyconnecttimespan timeout readpreference readpreference at mongodbdriverinternalmultipleinstancemongoserverproxychooseserverinstancereadpreference readpreference at mongodbdrivermongoserveracquireconnectionmongodatabase database readpreference readpreference at at at at source at mongodbdrivermongocollectionfindoneasimongoquery query at queryif we remove the replica set from the connection the problem goes away i saw another issue mentioning this problem where removing disconnect solves the issue we dont use disconnect this is a blocking issue

this line will throw an error if the query to countdocuments does not match codejava dbcollectionsomecollectioncountdocumentsfoo blah code uncaught typeerror cannot read property n of undefined docs is an empty array if the query does not match

update move repo modal on general settings page with dropdown so users can select an already existing repo also add validation to modal

a query matches all numbers greater than non numbers do not match should match the opposite values of which means all numbers and all non numbers i believe the matcher implements this behavior but the fieldrange constructor employs type bracketing and only matches numbers is a testcodec dbccdropcsave does not match printjson cfind a toarray printjson cfind a explain does match printjson cfind a not toarray printjson cfind a not explain censureindex does not match printjson cfind a toarray includes numbers printjson cfind a explain now does not match printjson cfind a not toarray includes numbers but should also include non numbers printjson cfind a not explain codewithout an index a not gt query will return docs where the queried field is missing adding an index on that field causes the query to return only documents where the field is present see example below mongomongodb shell version to test dbfoodroptrue dbfooinsertname a data dbfooinsertname b dbfoofinddata not gt id name a data id name b dbfooensureindexdata dbfoofinddata not gt id name a data

using gdb on solaris i was unable to access any threadlocal variables in spidermonkey noformat gdb bt in sorecv from in sorecv from in recv from in mongorecv thisthisentry bufbufentry at in mongounsaferecv buf this at mongorecv this bufbufentry at in mongorecv thisthisentry m at in mongocall this tosend response at in mongocall this tosend response assertok actualserver at in mongoruncommandwithmetadata this database command metadata commandargs at in mongoruncommandwithmetadataandtarget thisthisentry database command metadata commandargs at in mongoruncommandwithtarget this dbname cmd info options at in mongocall cx args at in bool mongowrapconstrainedmethodjscontext unsigned int jsvalue at in jscalljsnative args native cx at jsinvoke cx args construct at in interpret cx state at in jsrunscript cxcxentry state at in jsinvoke cxcxentry args constructconstructentryjsnoconstruct at in jsfuncall cxcxentry argc vp at in jscalljsnative args native cx at jsinvoke cx args construct at in interpret cx state at in jsrunscript cxcxentry state at in jsinvoke cxcxentry args constructconstructentryjsnoconstruct at in jsfunapply cxcxentry argc vp at in jscalljsnative args native cx at jsinvoke cx args construct at in interpret cx state at in jsrunscript cxcxentry state at in jsinvoke cxcxentry args constructconstructentryjsnoconstruct at in jsinvoke cx thisv fval argv rval at in jscallfunctionvalue cx obj objentry fval fvalentry args rval rvalentry at in jscall rval args fun thisobj cx at mongocallthreadargs this args at in mongorunvoid at in nsprthreadroutine arg at in executenativethreadroutine in thrpsetup from in from in gdb p mongokcurrentscope cannot find threadlocal variables on this target noformat

the primary server on my replicaset became slow due to a runaway process on the server eventually it failed altogether and a new primary was electedmy client applications failed to switch to the new primaryfirst of all i got thismongodbdrivermongoconnectionexception unable to connect to the primary member of the replica set timeout waiting for a mongoconnection systemtimeoutexception timeout waiting for a mongoconnection at mongodbdriverinternalmongoconnectionpoolacquireconnection mongodbdrivermongodatabase database in at mongodbdrivermongoserverinstanceconnect boolean slaveok in end of inner exception stack trace at mongodbdriverinternalreplicasetconnectorconnect timespan timeout connectwaitfor waitfor in at mongodbdrivermongoserverconnect timespan timeout connectwaitfor waitfor in at mongodbdrivermongoserverconnect connectwaitfor waitfor in at mongodbdrivermongoserverchooseserverinstance boolean slaveok in at mongodbdrivermongoserveracquireconnection mongodbdrivermongodatabase database boolean slaveok in at mongodbdrivermongocollectioninsertbatch systemtype nominaltype ienumerable documents mongodbdrivermongoinsertoptions options in at mongodbdrivermongocollectioninsert systemtype nominaltype systemobject document mongodbdrivermongoinsertoptions options in at mongodbdrivermongocollectioninsert systemtype nominaltype systemobject document mongodbdriversafemode safemode in at mongodbdrivermongocollectioninsert systemtype nominaltype systemobject document in at mongodbdrivermongocollectioninsert mongodbbsonbsondocument document in at msgexchangemessagessvraccountsaccountupdategps msgexchangemessagesgpstrackermessage trackermessage in and then this even after the server problem was fixed and the primary was switched back to the original serversystemtimeoutexception timeout waiting for a mongoconnection at mongodbdriverinternalmongoconnectionpoolacquireconnection mongodbdrivermongodatabase database in at mongodbdrivermongoserverinstanceacquireconnection mongodbdrivermongodatabase database in at mongodbdrivermongoserveracquireconnection mongodbdrivermongodatabase database boolean slaveok in at in at in at in at msgservercollectionsaccountsclsstoreforwardutilssendstoreforward msgexchangemessagessvraccountsaccount account in i had to restart my client applications

for support on os x we will need to use clang since os x gcc is very old however clang support on os x basically requires that we use libc rather than the also very old libstdc we will need to educate scons about invoking clang with the stdliblibc flag to support on os x

masterslave functionality is not compatible with changes being made to the sharding infrastructure going forward as such masterslave functionality may not be supported in sharded environments in the future we would like to deprecate this functionality as of right now so as to give users an opportunity to find alternatives and reach out to us to discuss their use cases such that we can make sure they are fully covered before removing the functionality