| Summary: | System becomes unresponsive under memory pressure | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise Linux 5 | Reporter: | Casey Dahlin <cdahlin> | ||||
| Component: | kernel | Assignee: | Johannes Weiner <jweiner> | ||||
| Status: | CLOSED INSUFFICIENT_DATA | QA Contact: | Zhouping Liu <zliu> | ||||
| Severity: | urgent | Docs Contact: | |||||
| Priority: | urgent | ||||||
| Version: | 5.6 | CC: | ccui, jmarchan, lwang, lwoodman, nobody+295318, vanhoof | ||||
| Target Milestone: | rc | ||||||
| Target Release: | --- | ||||||
| Hardware: | Unspecified | ||||||
| OS: | Unspecified | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | Bug Fix | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2012-04-11 17:34:18 UTC | Type: | --- | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Bug Depends On: | |||||||
| Bug Blocks: | 743405, 807974 | ||||||
| Attachments: |
|
||||||
This request was evaluated by Red Hat Product Management for inclusion in Red Hat Enterprise Linux 5.7 and Red Hat does not plan to fix this issue the currently developed update. Contact your manager or support representative in case you need to escalate this bug. In RHEL5 there is no sync IO from reclaim, that came in with lumpy reclaim for 2.6.24. Is it possible to get a better description of what's being observed? What's the value of /proc/sys/vm/dirty_ratio? And does lowering it to 20 help the workload? Customer is gone. Haven't heard of this being reproduced anywhere else. Going to go ahead and close. |
Created attachment 486893 [details] Patch to raise bar for synchronous IO in pageout The kernel tends to begin synchronous IO a bit too early in the reclaim path, which can result in lots of IO congestion and severe unresponsiveness. The attached patch makes the kernel a bit less eager to move to synchronous IO when larger allocations are being made, preventing this condition.