| [00:01] <flaccid> :) | |
| [00:04] <dialtone> at this point I just want to disable it and redirect stdout and stderr to a file | |
| [00:04] <dialtone> without having to repackage the instance | |
| [09:31] <progre55> hi guys. I'm trying to run an instance, but can not connect to it. the console output shows "waiting for metadata service at http://169.254.169.254/2009-04-04/meta-data/instance-id 08:19:06 [ 1/100]: url error [timed out]" | |
| [09:31] <progre55> Any suggestions, please? | |
| [10:36] <progre55> hmm.. for some reason, all the instances I'm running have the same public-IP, although all of them are "running" at the same time, and the private IP changes to a valid-IP after a while with a message on nc.log saying "discovered private IP xx.xx.xx.xx for instance i-408F076D" | |
| [10:40] <progre55> oh and cloud-error.log says "ERROR Found 2 vms with the same address: com.eucalyptus.net.util.ClusterAddressInfo@e30c8f0c -> i-408F076D i-48A408AD" | |
| [10:42] * progre55 is starting to dislike eucalyptus and thining to go back to amazon AWS ) | |
| [14:09] <progre55> smoser: hi | |
| [14:09] <progre55> here? | |
| [14:10] <smoser> hi | |
| [14:11] <smoser> progre55, i'm not sure why you would get multiple public ips to ssame instance. | |
| [14:11] <progre55> sorry for disturbing, it's just, afaik you're the only one active in this channal =) | |
| [14:11] <smoser> obviously thats not good | |
| [14:11] <smoser> we've seen the metadata service broken-ness from time to time (there are bugs opened). | |
| [14:11] <progre55> smoser: well, I've already reinstalled all the machines ) | |
| [14:11] <smoser> well, most people here have "real work" to be doing (myself included) | |
| [14:11] <progre55> yeah I understand =) | |
| [14:11] <smoser> the one other place to go for support on eucalyptus is the eucalyptus forums | |
| [14:11] <progre55> just wanted to ask.. | |
| [14:12] <smoser> oh yeah | |
| [14:12] <smoser> no harm | |
| [14:12] <smoser> what does euca-describe-addresses say (as admin) . can you pastebin that ? | |
| [14:12] <smoser> while you have a couple instances running | |
| [14:12] <smoser> and also | |
| [14:12] <progre55> I've set up a machine as cc+nc+storage+walrus and another machine as nc | |
| [14:12] <smoser> euca_conf --list-nodes | |
| [14:12] <progre55> yeah, about that, list-nodes | |
| [14:12] <progre55> so I have 2 machines, meaning 2 nodes | |
| [14:13] <progre55> but list-nodes shows me 3 | |
| [14:13] <progre55> 2 of them being the main machine + the separate node | |
| [14:13] <progre55> but the main machine nodes have 2 different IP addresses, one valid and one invalid | |
| [14:13] <progre55> is that okay? | |
| [14:14] <smoser> hm. | |
| [14:14] <progre55> and although I have 8 cpu cores all together, euca-describe-availability-zones says I can run 10 instances with 1 cpu each.. | |
| [14:14] <smoser> i would ditch nc on the cc+storage+walrus machine | |
| [14:14] <smoser> that is not something so often tested. | |
| [14:14] <progre55> oh, I have 6 cpu cores, 4 from the cc and 2 from the nc machine | |
| [14:15] <smoser> most common test scenario was cc+storage+walrus and ncs | |
| [14:15] <progre55> hmm.. I see | |
| [14:15] <progre55> what if I have a 32bit 1 core machine for cc? | |
| [14:16] <progre55> and 64bit quadro machines for NCs | |
| [14:16] <progre55> would it work? | |
| [14:16] <progre55> it's just, I have only 2 powerful machines, and plus a bunch of crappy machines.. =) | |
| [14:16] <progre55> and I wanted to use the powerful machines as NCs | |
| === dendrobates is now known as dendro-afk | |
| === dendro-afk is now known as dendrobates | |
| [14:47] <smoser> yeah, i would want ot use powerfull for NC alos | |
| [14:47] <smoser> so i might try cc+storage+walrus on "crappy" | |
| [15:01] <progre55> smoser: thanks. but can cc be a 32 bit machine with no KVM? | |
| [15:01] <smoser> yes | |
| [15:02] <progre55> oh, awesome =) | |
| [15:02] <progre55> smoser: thanks | |
| === dendrobates is now known as dendro-afk | |
| === dendro-afk is now known as dendrobates | |
| === dendrobates is now known as dendro-afk | |
| === dendro-afk is now known as dendrobates | |
| === mathiaz_ is now known as mathiaz | |
| === dendrobates is now known as dendro-afk | |
| === dendro-afk is now known as dendrobates | |
| [18:48] <hallyn> smoser: say, to run ec2-describe-images on stock lucid server, should that work with just ec2-ami-tools, ec2-api-tools, and sun-java6-jre installed? bc i'm getting not-found java classes, but woudl prefer not to install the .zips from amazon.com | |
| [18:50] <smoser> hallyn, yes. you should just be able to apt-get install ec2-api-tools from multiverse | |
| [18:50] <hallyn> <grimace> | |
| [18:51] <smoser> it does work.. | |
| [18:51] <smoser> what errors are you getting ? | |
| [18:53] <hallyn> smoser: http://pastebin.com/4VsiWEtr | |
| [18:54] <smoser> gnu.gcj.runtime.SystemClassLoader | |
| [18:54] <smoser> java is gcj | |
| [18:55] <hallyn> i wonder where that is specified | |
| [18:55] <hallyn> bc i dont' have gcj | |
| [18:55] <smoser> update-alternatives --query java | |
| [18:56] <hallyn> Best: /usr/bin/gij-4.4 | |
| [18:58] <hallyn> smoser: thx, i'm in business. | |
| === dendrobates is now known as dendro-afk | |
| === dendro-afk is now known as dendrobates | |
| === kim0 is now known as kim0_away | |