I'm not sure that this is a Proxmox question at all, perhaps a Linux question but I'll ask anyway.
I've got 2x HP Proliant servers both purchased in November and with identical spec, set up as a cluster, both version 1.4. One of them is running slowly. The problem seems to be specific to the network, and basically what happens is it freezes and then starts again after a short while. The problem can be reproduced just by typing on the keyboard at the shell repeatedly, where it will lock up and then release. If I type the same gibberish at the actual machine rather than via putty, I don't get the slowness. I've also swapped network leads with the proxmox master, and so can rule out the network infrastructure as a source of the problem.
I noticed this when one of the vm's had a very slow speed running FOG (http://fogproject.org) which I am using for hard disk imaging. I was getting a effective speed of about 70-100 MB a minute rather than 1+ GB per minute on the 100MB network I would normally expect from this (decompressing hard disk image) task.
There are no errors when it comes to the cluster, the master is fine with it. I am migrating both hosts now.
I also checked out hddtemp which came out at a cool 32C so can probably eliminate overheating as a cause.
Could anybody recommend any tools or techniques to troubleshoot this, or do they have any ideas as to what might be causing it?
I've got 2x HP Proliant servers both purchased in November and with identical spec, set up as a cluster, both version 1.4. One of them is running slowly. The problem seems to be specific to the network, and basically what happens is it freezes and then starts again after a short while. The problem can be reproduced just by typing on the keyboard at the shell repeatedly, where it will lock up and then release. If I type the same gibberish at the actual machine rather than via putty, I don't get the slowness. I've also swapped network leads with the proxmox master, and so can rule out the network infrastructure as a source of the problem.
I noticed this when one of the vm's had a very slow speed running FOG (http://fogproject.org) which I am using for hard disk imaging. I was getting a effective speed of about 70-100 MB a minute rather than 1+ GB per minute on the 100MB network I would normally expect from this (decompressing hard disk image) task.
There are no errors when it comes to the cluster, the master is fine with it. I am migrating both hosts now.
I also checked out hddtemp which came out at a cool 32C so can probably eliminate overheating as a cause.
Could anybody recommend any tools or techniques to troubleshoot this, or do they have any ideas as to what might be causing it?