Thanks a lot! Will try it. Sent from my iPhone
On Sep 28, 2011, at 12:03 AM, Peter Schuller <peter.schul...@infidyne.com> wrote: >> again.... I was doing repair on single CF and it crashed because of OOM, >> leaving 286GB data(should be 40GB). the problem here is it take very very >> long to make the node back to alive, seems because of it was loading row >> cache. the last time I encountered this, I did people suggested that delete >> everything in saved_cache directory and update the schema to set row/key >> cache to 0. but it seems is cluster wide and affect other nodes. >> so is there anyway to stop the node to load cache while starting? > > Just removing the saved_cache stuff on the node accomplishes that. > > You can also change the cache sizes transiently on a single node by > using nodetool, but there is no need to do that if all you're after is > to avoid reading in row cache on startup. Simply removing the saved > caches is enough. > > -- > / Peter Schuller (@scode on twitter)
