From: marton.szabo on
hi,

we are testing several java applications on suse 10 with 16 cpu´s
(intel xeon 3.00ghz) and 16GB RAM on a ibm blade. sometimes it happens
that suse freezes randomly and i can´t access the machine via ssh or
anything else anymore. there is nothing useful in the log files.

the only hint i can see is with nmon (with option: verbose mode):
"Danger -> Top Disk sda %busy 100.3%"

you guys have any idea what this could mean?


thanks in advance, marton
From: houghi on
marton.szabo(a)ivz.de wrote:
> hi,
>
> we are testing several java applications on suse 10 with 16 cpu�s
> (intel xeon 3.00ghz) and 16GB RAM on a ibm blade. sometimes it happens
> that suse freezes randomly and i can�t access the machine via ssh or
> anything else anymore. there is nothing useful in the log files.
>
> the only hint i can see is with nmon (with option: verbose mode):
> "Danger -> Top Disk sda %busy 100.3%"
>
> you guys have any idea what this could mean?
>
>

Just guestimating. Probably sda holds tmp and/or swap and that is being
written to. First disable the HD cache and see what happens. Also if you
can figure out who or what is writing to it, you could reduce the amount
it can write and still leave some space for root to access the machine
(unless it is root)

houghi
--
This was written under the influence of the following:
| Artist : Evanescence
| Song : Imaginary
| Album : Anywhere But Home
From: marton.szabo on
> Just guestimating. Probably sda holds tmp and/or swap and that is being
> written to. First disable the HD cache and see what happens. Also if you
> can figure out who or what is writing to it, you could reduce the amount
> it can write and still leave some space for root to access the machine
> (unless it is root)


thank you for your answer. do you know where i can disable the HD
cache? i already was checking the bios but i couldn´t find anything
there.
maybe with hdparm?


thanks in advance, marton
From: Chris Cox on
marton.szabo(a)ivz.de wrote:
> hi,
>
> we are testing several java applications on suse 10 with 16 cpu�s
> (intel xeon 3.00ghz) and 16GB RAM on a ibm blade. sometimes it happens
> that suse freezes randomly and i can�t access the machine via ssh or
> anything else anymore. there is nothing useful in the log files.
>
> the only hint i can see is with nmon (with option: verbose mode):
> "Danger -> Top Disk sda %busy 100.3%"
>
> you guys have any idea what this could mean?

Not sure... however the ONLY thing I know that will lock up an
openSUSE 10 or SLES 10 box is the evil zmd. Disable the
novell zmd if were talking about SLES and see if you experience
anymore lock ups. Zmd simply runs amuck sometimes (and no... version
level of SLES doesn't make a difference).

Just do Online Updates periodically with YaST.