Afther of some weeks running a stable configuration I find the problem.<br><br>I using a crossover cable to interconect the servers ERROR!!!!!!!!!!!!!!!<br><br>After of buy a simple Netgear switch 5 port 1G the server was runing weeks with no problem.. and drbd too.<br>
<br>In some test I see in one server:<br>[ 3941.506695] (518,1):ocfs2_query_inode_wipe:906 ERROR: Inode 6970563 (on-disk 6970563) not orphaned! Disk flags 0x1, inode flags 0x0<br>[ 3941.506724] (518,1):ocfs2_delete_inode:1034 ERROR: status = -17<br>
<br>But no more problems I hope that is not a hidden problem. :)<br><br>regards!!!<br><br><div class="gmail_quote">On Mon, Feb 13, 2012 at 6:54 PM, Eduardo Diaz - Gmail <span dir="ltr"><<a href="mailto:ediazrod@gmail.com">ediazrod@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Serius?? I make many test in 32 bits environment and don't has any problem.....<br><br>I will take another approach and I will put a switch in the interconnect.. (now I has a crossover cable)..<br>
<br>If still have problems I will look a view of of migrating to 64bits.. :-)<br>
<br>Thanks for the tip :)<div class="HOEnZb"><div class="h5"><br><br><br><div class="gmail_quote">On Fri, Feb 10, 2012 at 1:54 PM, Adi Kriegisch <span dir="ltr"><<a href="mailto:adi@cg.tuwien.ac.at" target="_blank">adi@cg.tuwien.ac.at</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Dear Eduardo,<br>
<div><br>
> I shutdown the filesystem an make a fsck.ocfs2 and there is many<br>
> errors y cluster file but there is no way to test that the ocfs2 are<br>
> ok? I can stop in night but for me this are crazy, because every to<br>
> months the filesystem are broken and if I stop one node the running<br>
> node go down...<br>
><br>
> I have all system in debian squezee with ocfs2 1.6.3<br>
><br>
> Any Ideas??<br>
</div>We had similar issues (also running debian squeeze 32bit). At the time we<br>
suspected having not enough LOWMEM available for the recovery to complete<br>
successfully.<br>
Switching to amd64 solved the issue for us. Luckily ocfs2 is able to run<br>
with mixed 32bit and 64bit clients so we could migrate our servers one by<br>
one without too much interrupting production.<br>
<span><font color="#888888"><br>
-- Adi<br>
</font></span></blockquote></div><br>
</div></div></blockquote></div><br>