The subject may say it all, but I was wondering... We've had several occasions where AIX gets a high blocked queue under the following conditions. This is probably obvious, but I have to ask anyway. AIX 4.3.3/Samba 2.0.7 12 Filesystems available as Samba shares *and* exported vis NFS (I know, I know...dumbass) to another AIX 4.3.3 box. Here's what happens...with oplocks on AIX will eventually throw-up via what appears to be a deadlock between Samba and the local nfsd. When I turned oplocks off, the problem has not happened (yet, anyway). I'm sure this is the fix, although not ideal and since a terabyte of storage is $400,000(US), I don't see us duplicating the storage area...so my questions are: 1) How much of a performance hit can I expect without oplocks? 2) How bad is it to rely only on NFS as a locking mechanism if lockd and statd are running on both systems? Bandwidth is not an issue at our site, we have bandwidth to burn, I'm just wondering what kind of problem I'm creating for myself. I'm assuming the NFS lock mechanism must be kernel based and should be reliable for my needs. Could anyone shed some light on why I might not want to do this? Thanks, Bill