[LU-5545] a lot of warnings like ptlrpc_at_adj_net_latency() Reported service time 7 > total measured time 0 Created: 26/Aug/14 Updated: 27/Apr/15 Resolved: 06/Nov/14 |
|
| Status: | Resolved |
| Project: | Lustre |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | Lustre 2.7.0, Lustre 2.5.4 |
| Type: | Bug | Priority: | Minor |
| Reporter: | Liang Zhen (Inactive) | Assignee: | Liang Zhen (Inactive) |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | None | ||
| Severity: | 3 |
| Rank (Obsolete): | 15448 |
| Description |
|
when we enable message dropping on routers, and run a simple MPI program to repeat fstat and fchmod, I saw a lot of warnings like this in log: Lustre: 17660:0:(client.c:304:ptlrpc_at_adj_net_latency()) Reported service time 7 > total measured time 0 Lustre: 17663:0:(client.c:304:ptlrpc_at_adj_net_latency()) Reported service time 7 > total measured time 0 they may be harmless, but it's better to fix them. |
| Comments |
| Comment by Chris Horn [ 28/Aug/14 ] |
|
When client re-sends a request, the server will drop it if it is already in progress. Client then is measuring the service time from the sent time of the resend, while the server will report service time of original request. This may be the cause of the disparity. |
| Comment by Li Wei (Inactive) [ 29/Aug/14 ] |
|
Chris, yes, that's exactly what I saw. |
| Comment by Liang Zhen (Inactive) [ 23/Sep/14 ] |
|
Thanks Chris, patch is here: http://review.whamcloud.com/12018 |
| Comment by Liang Zhen (Inactive) [ 06/Nov/14 ] |
|
patch landed |
| Comment by Gerrit Updater [ 25/Nov/14 ] |
|
Jian Yu (jian.yu@intel.com) uploaded a new patch: http://review.whamcloud.com/12855 |
| Comment by Gerrit Updater [ 04/Dec/14 ] |
|
Oleg Drokin (oleg.drokin@intel.com) merged in patch http://review.whamcloud.com/12855/ |