I have the following NFS mount:
server:path /mnt/path nfs rw,hard,intr,nfsvers=4,async 0 0
Pings to the server are ~60ms, which is quite reasonable, yet any access to the mount is very slow. Copying a small image takes 10-30 seconds, and even ls in small directories takes several seconds. This is true for both NFSv3 and NFSv4. This occurs when ~10 writes/second are being attempted. With no load, access/write times are shorter, but still nowhere near immediate.
This is strange to me, since even if there is a transfer bottleneck, I would expect async to allow my stat and write calls to return as quickly as the round trip takes, without blocking for so long.
This leads me to wonder if my mounts are actually honoring the async option. In /proc/mounts, the following is listed:
server:path /mnt/path nfs4 rw,relatime,vers=4.1,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,port=0,timeo=600,retrans=2,sec=sys,clientaddr=xxx.xxx.xxx.xxx,local_lock=none,addr=xxx.xxx.xxx.xxx 0 0
async is not listed, but neither is sync, and since async is supposed to be the default, it is unclear to me whether the option is actually active. Is there a way to tell?
