Hi NFS folks, I've noticed a pretty nasty regression in our NFS capability between 5.17 and 5.18-rc1. I've tried to bisect but not having any luck. The problem I'm seeing is it takes 3 minutes to copy a file from NFS to the local disk. When it should take less than half a second, which it did up through 5.17. It doesn't seem to be network related, but can't rule that out completely. I tried to bisect but the problem can be intermittent. Some runs I'll see a problem in 3 out of 100 cycles, sometimes 0 out of 100. Sometimes I'll see it 100 out of 100. The latest attempt I have is at 5.18-rc4 and here's what I see when I 1) create file with dd, 2) copy local disk to NFS mount, 3) copy NFS to local disk. Test 2 Creating /dev/shm/run_nfs_test.junk... 262144+0 records in 262144+0 records out 268435456 bytes (268 MB, 256 MiB) copied, 1.29037 s, 208 MB/s Done copy /dev/shm/run_nfs_test.junk to /mnt/nfs_test/run_nfs_test.junk... real 0m0.502s user 0m0.001s sys 0m0.250s Done copy /mnt/nfs_test/run_nfs_test.junk to /dev/shm/run_nfs_test.tmp... real 4m11.835s user 0m0.001s sys 0m0.277s Done Comparing files... Done Remove /dev/shm/run_nfs_test.tmp... real 0m0.031s user 0m0.000s sys 0m0.031s Done Remove /dev/shm/run_nfs_test.junk... real 0m0.031s user 0m0.001s sys 0m0.030s Done Remove /mnt/nfs_test/run_nfs_test.junk... real 0m0.024s user 0m0.001s sys 0m0.022s Done Create the server with a 4G RAM disk: mount -t tmpfs -o size=4096M tmpfs /mnt/nfs_test exportfs -o fsid=0,rw,async,insecure,no_root_squash 192.168.254.0/255.255.255.0:/mnt/nfs_test Mount on the client side with: mount -o rdma,port=20049 192.168.254.1:/mnt/nfs_test /mnt/nfs_test Any advice on tracking this down further would be greatly appreciated. Thanks -Denny
