Network Storage Protocols Discussions
if I do an svn co "repo" it takes ~4 minutes to check out. If I simply copy the files to the same volume it's just over a minute.
If I check out or copy to my workstation it's just over a minute for each so I don't think this is something on the server side but I am not an SVN guy.
If I run a stats show for nfs ops on that volume while the checkout is running I see between 0 and 200 IOPS of r/w but 3k (!) of "other".
Any suggestions? This is killing our build times. NFSv3, default mount options from a Ubuntu 9.10 Subversion server.
Those operation are likely file attributes requests. Are you running with attribute caching disabled for this NFS mount?
On the host or the client? I am not specifying anything n the export except an admin host. The client is using the following:
array02:/vol/engineering /net nfs rw,noatime,vers=3,rsize=65536,wsize=65536,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=10.0.17.214,mountvers=3,mountproto=tcp,addr=10.0.17.214 0 0