Gorak,
There are two parameters. The heartbeat_timeout has a limit of 60000,
but it is the heartbeat_quantum which I think is the one to tune having
taken a peak at the code. Did you adjust that too? If not, please try
that. Otherwise, I'm not sure if there is anything you can do.
Regards,
Tim
---
On 07/01/11 08:09, gorak wrote:
I actually increased it to 60000 (max limit) but still facing the panics. I'm
really surpised on what's causing these delays. I'm definitely not expecting a
production grade cluster on a laptop, but at the very least it should be alive,
and allow me to perform some minimal tasks.
Gorak,
You could try increasing heartbeat_quantum to say
2000 (2 sec) and
heartbeat_timeout to 20000 (20 sec). I'm not certain
that will make any
difference, but it is worth a try. I did have a quick
look at the code
and from what I think I understood, the quantum is
taken into account in
the delay processing. (I may be wrong!)
Regards,
Tim
---
On 06/28/11 23:40, gorak wrote:
Hmm....Tried almost everything on that document,
but still getting the pm_tick delay panics atleast
once every hour. It's becoming a real challenge to
have the cluster up and running with this panic. I
have a pretty good system with i7 processor and 8gb
of memory, but not sure why this issue is still
happening. Any ideas?
Hi,
have a look at page 14/15 within the white paper
published quite some
time back at
http://blogs.oracle.com/TF/entry/new_white_paper_pract
icing_solaris
It shows you how to change syslog.conf to not
catch
the pm_tick delay
messages - which makes syslogd and then the whole
virtual system very
unresponsive, which then might lead to the crash
you
see.
Regards
Thorsten
rote:
I must really convey my sincere thanks to you for
pointing that out. That fixed the problem for me.
It's not enough to just turn global fencing off
during scinstall. We need to turn it off for
individual quorum disks too. I did it and that
fixed
the problem. I just can't believe it fixed the
problem. I'd had been struggling with this for
over a
week now. Thanks frueauf. Just to mention I still
get
errors like "CMM: Issuing a NULL preempt failed on
quorum device...". I believe it's nothing to
worry.
Now on to another issue, that's annoying too.
Anyone pls help me if you can. I have frequent
cluster panics because of pm_tick delays. I've set
the pm_tick threshold to the maximum limit
allowed,
but i still get pm_tick panics. Any ideas on it?
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~~
Sitz der Gesellschaft:
ORACLE Deutschland B.V.& Co. KG
Hauptverwaltung: Riesstr. 25, D-80992 Muenchen
Registergericht: Amtsgericht Muenchen, HRA 95603
Komplementaerin: ORACLE Deutschland Verwaltung
B.V.
Hertogswetering 163/167, 3543 AS Utrecht,
Niederlande
Handelsregister der Handelskammer
Midden-Niederlande, Nr. 30143697
Geschaeftsfuehrer: Juergen Kunz, Marcel van de
Molen, Alexander van
er Ven
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~~~~
_______________________________________________
ha-clusters-discuss mailing list
ha-clusters-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/ha-cluste
rs-discuss
--
Tim Read
Software Developer
Solaris Availability Engineering
Oracle Corporation UK Ltd
Springfield
Linlithgow
EH49 7LR
Phone: +44 (0)1506 672 684
Mobile: +44 (0)7802 212 137
Twitter: @timread
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~
NOTICE: This email message is for the sole use of the
intended
recipient(s) and may contain confidential and
privileged information.
Any unauthorized review, use, disclosure or
distribution is prohibited.
If you are not the intended recipient, please contact
the sender by
reply email and destroy all copies of the original
message.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
~~~~~~~~~~~~~~~~~~~
_______________________________________________
ha-clusters-discuss mailing list
ha-clusters-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/ha-cluste
rs-discuss
--
Tim Read
Software Developer
Solaris Availability Engineering
Oracle Corporation UK Ltd
Springfield
Linlithgow
EH49 7LR
Phone: +44 (0)1506 672 684
Mobile: +44 (0)7802 212 137
Twitter: @timread
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
NOTICE: This email message is for the sole use of the intended
recipient(s) and may contain confidential and privileged information.
Any unauthorized review, use, disclosure or distribution is prohibited.
If you are not the intended recipient, please contact the sender by
reply email and destroy all copies of the original message.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
_______________________________________________
ha-clusters-discuss mailing list
ha-clusters-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/ha-clusters-discuss
