Hi, When I run few "hard-writers" I get problems with my 15 realtime processes (do some very small writes, just 2 pages per-time): they start thrashing. I thought it's caused by write-back and was waiting for Greg Tellen's per-cgroup dirty page accounting patch. Before testing it I tried to change threshold in page-writeback.c:get_dirty_limits(), I set dirty_ratio for RT process 80% (instead of just extra dirty / 4), but it didn't help me. What else can cause problems? I'm linux kernel newbie and will appreciate any addvices. -- Evgeniy Ivanov -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxxx For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom policy in Canada: sign http://dissolvethecrtc.ca/ Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>