[17:45] <axolx> hi -- anyone experiencing connecitivty problems with ec2, us-east-1b?
[17:46] <erichammond> axolx: There have been recent issues: http://ec2status.notlong.com
[17:46] <axolx> erichammond: yes but those notes claim they have been fixed as of 1 hour ago.
[17:48] <erichammond> axolx: Yep, add your traceroutes to http://developer.amazonwebservices.com/connect/thread.jspa?threadID=36409&tstart=0
[17:48] <erichammond> axolx: Note that availability zone names do not apply across accounts: http://www.google.com/search?hl=en&q=alestic%20availability
[17:48] <erichammond> er, http://alestic.com/2009/07/ec2-availability-zones
[17:49] <axolx> erichammond: this: http://developer.amazonwebservices.com/connect/thread.jspa?threadID=36409&tstart=0 is not loading either. oh my
[20:50] <smoser> wonder if anyone else has seen something similar to this
[20:51] <smoser> i started a c1.medium today rather than a c1.xlarge for a task that is highly bottlenecked on 'gunzip'
[20:51] <smoser> i had noticed that according to http://docs.amazonwebservices.com/AWSEC2/latest/DeveloperGuide/index.html?instance-types.html
[20:52] <smoser> the medium had 2x2.5 CPU Units and the xlarge had 8x2.5 CPU units
[20:52] <smoser> since gunzip is not threaded i figured medium would be the same
[20:52] <smoser> not at all
[20:52] <smoser> after maybe 20 minutes, i decided to start a xlarge and see how they compared.
[20:53] <smoser> the xlarge has made up the difference in gunzipped filesize in maybe 5 minutes.
[20:53] <smoser> what i'm getting at is that not all CPU Units are created equal :)
[20:53] <smoser> ymmv
[20:56] <soren> Interesting.
[20:56] <smoser> and note, that each system has 2 cpus at 100% utilization
[20:57] <smoser> so its not as if disk was bottlenecking the medium
[20:57] <smoser> neither shows any usage of swap
[21:13] <pfibiger> hey all, i'm rebundling an alestic ami using these instructions: http://alestic.com/2009/06/ec2-ami-bundle ...the original ami prints the ssh key fingerprint in the console, but my rebundled ami doesn't. i'm sure it's something simple i'm missing, does anyone know what step i'm missing?
[21:27]  * pfibiger realized what it is :P
[21:28] <erichammond> pfibiger: You can fix this by chmod +x /etc/init.d/ec2-ssh-host-key-gen before bundling, but make sure to chmod -x before rebooting.
[21:29] <erichammond> An upcoming release of the AMIs will default to regenerating ssh keys on rebundled AMIs since you are the fourth person to think it should work this way and nobody has defended my original choice.
[21:29] <pfibiger> erichammond: yeah, i just realized that a minute or two ago. thanks for the amis and the rebundling instructions :)
[21:31] <erichammond> You could also install the updated boot script which works the way you want: http://ec2ubuntu.googlecode.com/svn/trunk/etc/init.d/ec2-ssh-host-key-gen
[21:32] <erichammond> smoser: That CPU test does not match my expectations either.  Deserves further investigation for those of us who live in this environment.
=== zul_ is now known as zul