On 07/07/2010 05:43 AM, Mel Gorman wrote:
How do you suggest tuning this? The modification I tried was "if N dirty pages are found during a SWAP_CLUSTER_MAX scan of pages, assume an average dirtying density of at least that during the time those pages were inserted on the LRU. In response, ask the flushers to flush 1.5X". This roughly responds to the conditions it finds as they are encountered and is based on scanning rates instead of time. It seemed like a reasonable option.
Your idea sounds like something we need to have, regardless of whether or not we fix the flusher to flush older inodes first (we probably should do that, too). I believe this for the simple reason that we could have too many dirty pages in one memory zone, while the flusher's dirty threshold is system wide. If we both fix the flusher to flush old inodes first and kick the flusher from the reclaim code, we should be golden. -- All rights reversed -- 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/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>