<html>
<head>
<meta content="text/html; charset=utf-8" http-equiv="Content-Type">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">Am 18.02.16 um 21:57 schrieb Schweiss,
Chip:<br>
</div>
<blockquote
cite="mid:CALeZrrQXDkqSXvQcD_55R2dxypx3Wu5jMaa2+awERX5gHdBXjg@mail.gmail.com"
type="cite">
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<div dir="ltr"><br>
<div class="gmail_extra"><br>
<div class="gmail_quote">On Thu, Feb 18, 2016 at 5:14 AM,
Michael Rasmussen <span dir="ltr"><<a
moz-do-not-send="true" href="mailto:mir@miras.org"
target="_blank">mir@miras.org</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex"><span
class="">On Thu, 18 Feb 2016 07:13:36 +0100<br>
Stephan Budach <<a moz-do-not-send="true"
href="mailto:stephan.budach@JVM.DE">stephan.budach@JVM.DE</a>>
wrote:<br>
<br>
><br>
> So, when I issue a simple ls -l on the folder of
the vdisks, while the switchover is happening, the
command somtimes comcludes in 18 to 20 seconds, but
sometime ls will just sit there for minutes.<br>
><br>
</span>This is a known limitation in NFS. NFS was never
intended to be<br>
clustered so what you experience is the NFS process on the
client side<br>
keeps kernel locks for the now unavailable NFS server and
any request<br>
to the process hangs waiting for these locks to be
resolved. This can<br>
be compared to a situation where you hot-swap a drive in
the pool<br>
without notifying the pool.<br>
<br>
Only way to resolve this is to forcefully kill all NFS
client processes<br>
and the restart the NFS client.<br>
<br>
</blockquote>
<div><br>
</div>
<div>I've been running RSF-1 on OmniOS since about r151008.
All my clients have always been NFSv3 and NFSv4. <br>
<br>
My memory is a bit fuzzy, but when I first started testing
RSF-1, OmniOS still had the Sun lock manager which was
later replaced with the BSD lock manager. This has had
many difficulties.<br>
<br>
</div>
<div>I do remember that fail overs when I first started with
RSF-1 never had these stalls, I believe this was because
the lock state was stored in the pool and the server
taking over the pool would inherit that state too. That
state is now lost when a pool is imported with the BSD
lock manager. <br>
<br>
</div>
<div>When I did testing I would do both full speed reading
and writing to the pool and force fail overs, both by
command line and by killing power on the active server.
Never did I have a fail over take more than about 30
seconds for NFS to fully resume data flow. <br>
<br>
</div>
<div>Others who know more about the BSD lock manager vs the
old Sun lock manager may be able to tell us more. I'd
also be curious if Nexenta has addressed this.<br>
<br>
</div>
<div>-Chip<br>
</div>
</div>
</div>
</div>
</blockquote>
I actually don't know, if it's the lock manager or the nfsd itself,
that caused this, but as I bounced all of them after I failed the
ZPOOL over while hammering it with reads and writes, lockd would
also have been part of the processes that had been restarted. And
remeber, this only happend when failing from to and back one host
in a rather quick manner.<br>
<br>
Nevertheless, RSF-1 seems to be a solid solution and I will very
likely implement it across several OmniOS boxes.<br>
<br>
Cheers,<br>
Stephan<br>
</body>
</html>