Dear all, I am struggling with FS-cache to have a big local cache on SSD drives to speed up our NFS mounts from our filer. We have a text search engine that access large files (flat text files from a few MB to 100GB max) through these mounts. Wright now, it seems the SSD cache is "heating" up slower than I expected. Out of 300GB I can see exchanged on the network, only 11GB is use in the cache. (My test server has 32GB of RAM, SSD cache is 300GB). Though, when I just do a simple copy of the same files, I can see the cache grow accordingly to the file's size. I was wondering if this could come from my FScache's settings. I am using default settings : brun 10% bcull 7% bstop 3% frun 10% fcull 7% fstop 3% Also, here are my fscache stats : Cat /proc/fs/fscache/stats FS-Cache statistics Cookies: idx=300 dat=120 spc=0 Objects: alc=161 nal=0 avl=161 ded=45 ChkAux : non=0 ok=120 upd=0 obs=0 Pages : mrk=7875596 unc=7634731 Acquire: n=420 nul=0 noc=0 ok=420 nbf=0 oom=0 Lookups: n=161 neg=1 pos=160 crt=1 tmo=0 Updates: n=0 nul=0 run=0 Relinqs: n=155 nul=0 wcr=0 rtr=0 AttrChg: n=0 ok=0 nbf=0 oom=0 run=0 Allocs : n=0 ok=0 wt=0 nbf=0 int=0 Allocs : ops=0 owt=0 abt=0 Retrvls: n=3162223 ok=579541 wt=38 nod=303342 nbf=2279340 int=0 oom=0 Retrvls: ops=882883 owt=0 abt=0 Stores : n=2479793 ok=2479787 agn=0 nbf=6 oom=0 Stores : ops=440329 run=2919824 pgs=2479539 rxd=2479787 olm=0 VmScan : nos=7634477 gon=0 bsy=0 can=248 Ops : pend=0 run=1323212 enq=8412714 can=0 rej=1 Ops : dfr=668 rel=1323212 gc=668 CacheOp: alo=0 luo=0 luc=0 gro=0 CacheOp: upo=0 dro=0 pto=0 atc=0 syn=0 CacheOp: rap=0 ras=0 alp=0 als=0 wrp=0 ucp=0 dsp=0 I can see a high level of nbf what I foubnd they were 'number of retrivals requests rejected -ENOBUFS'. And unc : Pacges uncached. Is there some buffer tunning to be done on the system for nfs mounts? Does any one of you have some experience with this? Thanks in advance !! Kind regards, Yann Lievin. -- Linux-cachefs mailing list Linux-cachefs@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cachefs
