Messages in this thread | | | Date | Thu, 19 Jan 2006 20:28:44 -0500 | From | Dave Jones <> | Subject | Re: Out of Memory: Killed process 16498 (java). |
| |
On Thu, Jan 19, 2006 at 05:03:05PM -0800, Andrew Morton wrote: > This is the bypass-the-batching patch. It's a reasonable thing to do, but I'd > just do it unconditionally and remove the code which clears > ->all_unreclaimable from free_pages_bulk(), if possible. > > Has this patch been shown to have any effect? If so, what was it, and > under what conditions?
Larry originally came up with the patch for our RHEL4 2.6.9 kernel, after customers were hitting OOM under some heavy workload he can probably recall better than I can.
iirc it didn't solve the users OOM entirely, but it does make their workload run longer before the kill happens.
A little while later, after Fedora users started reporting oom kills, I forward-ported it to 2.6.14, and threw it out in an update. The Fedora user in the bug report https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=175173 who on x86-64 with 5GB saw zone_dma exhausted saw a similar result, delays the kill, but it does still happen.
Hmm.
Dave - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |