[OpenAFS] Re: Solaris 10 deadlock issue

Aaron Knister aaronk@umbc.edu
Tue, 14 Jun 2011 19:08:57 -0400


--001636284fe80b10be04a5b42029
Content-Type: text/plain; charset=ISO-8859-1

On Tue, Jun 14, 2011 at 6:08 PM, Andrew Deason <adeason@sinenomine.net>wrote:

> On Tue, 14 Jun 2011 17:56:44 -0400
> Aaron Knister <aaronk@umbc.edu> wrote:
>
> > The issue can be mitigated if the cache size is raised to the value of
> > roughly half of the physical memory in the given system. The issue
> > appeared somewhere between Solaris 10 "u8" and "u9."
>
> "Mitigated" as in, it goes away entirely, or is less likely to occur, or
> ... ?
>

It's significantly less likely to occur. I only had it occur twice as
compared to the 20+ times it did not occur and both times I was
concatenating large files in my AFS home directory to test with.


>
> > I've reproduced the problem using OpenAFS 1.4.14.1, 1.5.78 and
> > 1.6.0pre6 and a Solaris 10 "u8" system with all of the latest patches
> > applied.
>
> Do you mean u9? "between u8 and u9" I would take to mean that the issue
> does not exist on u8, so... Or do you mean it happens on u8 only when
> you've applied certain patches?
>

Stock u8 doesn't have the issue. There's a patch in there somewhere that
after applying causes the issue. Figuring out which patch would be a real
headache.


>
> Do you know what the cache size and chunk size was in these situations?
> (You can get these from 'cmdebug -cache', in the future)
>

bash-3.2# /usr/afsws/bin/cmdebug -cache -servers localhost
Chunk files:   1953
Stat caches:   2929
Data caches:   1953
Volume caches: 200
Chunk size:    262144
Cache size:    499968 kB
Set time:      no
Cache type:    memory

Thanks!


>
> --
> Andrew Deason
> adeason@sinenomine.net
>
> _______________________________________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info
>



-- 
Aaron Knister
Systems Administrator
Division of Information Technology
University of Maryland, Baltimore County
aaronk@umbc.edu

--001636284fe80b10be04a5b42029
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

<div class=3D"gmail_quote">On Tue, Jun 14, 2011 at 6:08 PM, Andrew Deason <=
span dir=3D"ltr">&lt;<a href=3D"mailto:adeason@sinenomine.net">adeason@sine=
nomine.net</a>&gt;</span> wrote:<br><blockquote class=3D"gmail_quote" style=
=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">

<div class=3D"im">On Tue, 14 Jun 2011 17:56:44 -0400<br>
Aaron Knister &lt;<a href=3D"mailto:aaronk@umbc.edu">aaronk@umbc.edu</a>&gt=
; wrote:<br>
<br>
&gt; The issue can be mitigated if the cache size is raised to the value of=
<br>
&gt; roughly half of the physical memory in the given system. The issue<br>
&gt; appeared somewhere between Solaris 10 &quot;u8&quot; and &quot;u9.&quo=
t;<br>
<br>
</div>&quot;Mitigated&quot; as in, it goes away entirely, or is less likely=
 to occur, or<br>
... ?<br></blockquote><div><br></div><div>It&#39;s significantly less likel=
y to occur. I only had it occur twice as compared to the 20+ times it did n=
ot occur and both times I was concatenating large files in my AFS home dire=
ctory to test with.</div>

<div>=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;=
border-left:1px #ccc solid;padding-left:1ex;">
<div class=3D"im"><br>
&gt; I&#39;ve reproduced the problem using OpenAFS 1.4.14.1, 1.5.78 and<br>
&gt; 1.6.0pre6 and a Solaris 10 &quot;u8&quot; system with all of the lates=
t patches<br>
&gt; applied.<br>
<br>
</div>Do you mean u9? &quot;between u8 and u9&quot; I would take to mean th=
at the issue<br>
does not exist on u8, so... Or do you mean it happens on u8 only when<br>
you&#39;ve applied certain patches?<br></blockquote><div><br></div><div>Sto=
ck u8 doesn&#39;t have the issue. There&#39;s a patch in there somewhere th=
at after applying causes the issue. Figuring out which patch would be a rea=
l headache.</div>

<div>=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;=
border-left:1px #ccc solid;padding-left:1ex;">
<br>
Do you know what the cache size and chunk size was in these situations?<br>
(You can get these from &#39;cmdebug -cache&#39;, in the future)<br></block=
quote><div><br></div><div><div>bash-3.2# /usr/afsws/bin/cmdebug -cache -ser=
vers localhost</div><div>Chunk files: =A0 1953</div><div>Stat caches: =A0 2=
929</div>

<div>Data caches: =A0 1953</div><div>Volume caches: 200</div><div>Chunk siz=
e: =A0 =A0262144</div><div>Cache size: =A0 =A0499968 kB</div><div>Set time:=
 =A0 =A0 =A0no</div><div>Cache type: =A0 =A0memory</div></div><div><br></di=
v><div>Thanks!</div>

<div>=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;=
border-left:1px #ccc solid;padding-left:1ex;">
<font color=3D"#888888"><br>
--<br>
Andrew Deason<br>
<a href=3D"mailto:adeason@sinenomine.net">adeason@sinenomine.net</a><br>
<br>
_______________________________________________<br>
OpenAFS-info mailing list<br>
<a href=3D"mailto:OpenAFS-info@openafs.org">OpenAFS-info@openafs.org</a><br=
>
<a href=3D"https://lists.openafs.org/mailman/listinfo/openafs-info" target=
=3D"_blank">https://lists.openafs.org/mailman/listinfo/openafs-info</a><br>
</font></blockquote></div><br><br clear=3D"all"><br>-- <br>Aaron Knister<br=
>Systems Administrator<br>Division of Information Technology<br>University =
of Maryland, Baltimore County<br><a href=3D"mailto:aaronk@umbc.edu" target=
=3D"_blank">aaronk@umbc.edu</a><br>



--001636284fe80b10be04a5b42029--