Red Squirrel
No Lifer
- May 24, 2003
- 64,675
- 10,760
- 126
Still early to call it... but setting all shares to async may have actually fixed this problem. I was reluctant to use it because it means stuff could corrupt if a system crashes mid write, but if it fixes it then so be it. At least my sanity will be restored. This was just starting to drive me insane, it seemed to be getting worse to the point it was affecting the performance of my workstation. I don't know what's with operating systems, they all do this, but if a mapped drive has performance issues, it causes the entire system to lock up. Windows does it, and so does Linux. Even if you're not actually trying to access that drive.
Either way I'm getting over 100MB/sec on NFS now. I've never seen this on this network, ever. I was very lucky to get 10MB/sec most days, and it would slowly taper off during a transfer, until it stabilizes around 1KB/sec. Now I get a bit over 100MB/sec sustained. That is on par with what I should expect from a gig network. 125MB/sec absolute max.
Still not sure if the vms are running in async mode or not though, not sure if this is something it just picks up over time or if it would require a full remapping of the luns, which I'm actually not sure how to do live.
Either way I'm getting over 100MB/sec on NFS now. I've never seen this on this network, ever. I was very lucky to get 10MB/sec most days, and it would slowly taper off during a transfer, until it stabilizes around 1KB/sec. Now I get a bit over 100MB/sec sustained. That is on par with what I should expect from a gig network. 125MB/sec absolute max.
Still not sure if the vms are running in async mode or not though, not sure if this is something it just picks up over time or if it would require a full remapping of the luns, which I'm actually not sure how to do live.