Actually, it does work, one of the servers wasn't mounted. Here's the funny thing, my NFS share failed a few of the tests, and I have locd running and everything. The Gluster mounts had no problem: TEST : TRY TO WRITE ON A READ LOCK:.==================================================================================================.== TEST : TRY TO WRITE ON A WRITE LOCK:==================================================================================================== TEST : TRY TO READ ON A READ LOCK:==================================================================================================== TEST : TRY TO READ ON A WRITE LOCK:==================================================================================================== TEST : TRY TO SET A READ LOCK ON A READ LOCK:==================================================================================================== TEST : TRY TO SET A WRITE LOCK ON A WRITE LOCK:==================================================================================================== TEST : TRY TO SET A WRITE LOCK ON A READ LOCK:==================================================================================================== TEST : TRY TO SET A READ LOCK ON A WRITE LOCK:==================================================================================================== TEST : TRY TO READ LOCK THE WHOLE FILE BYTE BY BYTE:==================================================================================================== TEST : TRY TO WRITE LOCK THE WHOLE FILE BYTE BY BYTE:==================================================================================================== process number : 100 - Remote clients: 2 local client 1 - Total client 3 - Total concurent tests: 300 process number running test successfully : 300 process of 300 successfully ran test : WRITE ON A READ LOCK 300 process of 300 successfully ran test : WRITE ON A WRITE LOCK 300 process of 300 successfully ran test : READ ON A READ LOCK 300 process of 300 successfully ran test : READ ON A WRITE LOCK 300 process of 300 successfully ran test : SET A READ LOCK ON A READ LOCK 0 process of 300 successfully ran test : SET A WRITE LOCK ON A WRITE LOCK 0 process of 300 successfully ran test : SET A WRITE LOCK ON A READ LOCK 0 process of 300 successfully ran test : SET A READ LOCK ON A WRITE LOCK 300 process of 300 successfully ran test : READ LOCK THE WHOLE FILE BYTE BY BYTE 300 process of 300 successfully ran test : WRITE LOCK THE WHOLE FILE BYTE BY BYTE > Brian Taber wrote: >> I just ran lock tests I found in the QA section on 2 servers at the same >> time with 500 threads each, they all ran successfully. I am currently >> running the TLA version of gluster, I don't know if that makes a >> difference... > > So do I. Could you try the test in network mode? The source available > from the link on the GlusterFS QA page doesn't support network-based > tests (however, it does by documentation). There's an updated version at > > http://nfsv4.bullopensource.org/tools/tests/page45.php > > that can really do the trick. > > After compiling, you run > > locktests -f <file> \ > -n <number of processes per node> \ > -c <number of nodes - 1> > > on one server and > > locktests --server <servernode> > > on the other(s). This will test locking the same file from two clients. > > I would be interested in your results and your configuration, if you can > make them public. > > Thanks, > -- > cc > >>> Brian Taber wrote: >>>> I have been running Dovecot and Postfix with a 2 brick setup and posix >>>> locking enabled on each brick for a while with no problems. This is a >>>> Maildir setup... Dovecot is setup to use fcntl for locking.... >>>> remember >>>> that gluster does not support flock as fuse does not support it... >>> Yes, I'm aware of this. But in your place, I would stop doing this >>> until >>> the GlusterFS QA test runs successfully. If it doesn't, then not losing >>> data with such a setup is just a matter of luck. And for this storage >>> application, I can't accept the risk. >>> >>>>> On Fri, 30 Nov 2007, Székelyi Szabolcs wrote: >>>>> >>>>>> Everything works fine as long as I don't introduce locking, which is >>>>>> essential if one wishes to use the storage eg. as a backend for a >>>>>> mail >>>>>> server. >>>>> I'm sorry that I cannot answer your question, but do you have a >>>>> choice >>>>> in >>>>> what mail server to use? Those that use the Maildir format (Like for >>>>> example Postfix together with Courier-IMAP) does not need locking. > > > _______________________________________________ > Gluster-devel mailing list > Gluster-devel@xxxxxxxxxx > http://lists.nongnu.org/mailman/listinfo/gluster-devel >
