Re: Do we have a CPUTYPE=native and/or generic stability problem?

From: Olivier Smedts <olivier_at_gid0.org>
Date: Mon, 5 Nov 2012 10:51:10 +0100
Le lundi 5 novembre 2012, Olivier Smedts a écrit :

>
> Le lundi 5 novembre 2012, Olivier Smedts a écrit :
>
>>
>> Le dimanche 4 novembre 2012, Alexander Leidinger a écrit :
>
>
>>
>> The machine has 12 MB RAM (no swap configured), after nearly a day
>>> uptime it looks like this:
>>>
>>> ---snip---
>>> Mem: 348M Active, 599M Inact, 9281M Wired, 264K Cache, 1548M Free
>>> ARC: 7117M Total, 1607M MRU, 3996M MFU, 934K Anon, 208M Header, 1307M
>>> Other
>>> ---snip---
>>>
>>> I would not expect an internal compiler error when I run out of RAM.
>>
>> Not related, but I think you should configure
>>
> ...at least a little swap (sorry, damn "smart"-phone). When under memory
> pressure, ZFS sometimes don't have time to evict ARC memory and some MBs
> are written to swap instead of waiting. No real technical details
>
...here, I'll leave it to ZFS or VM gurus, but I always have few dozens of
MBs of my swap used even if I have lots of memory. And I promise my next
reply will be made on a real computer, not on a too small touch interface
with an editor which tries to replace all your english words by some of
your mother tongue words. Sorry for triple posting.


-- 
Olivier Smedts                                                 _
                                        ASCII ribbon campaign ( )
e-mail: olivier_at_gid0.org        - against HTML email & vCards  X
www: http://www.gid0.org    - against proprietary attachments / \

  "Il y a seulement 10 sortes de gens dans le monde :
  ceux qui comprennent le binaire,
  et ceux qui ne le comprennent pas."
Received on Mon Nov 05 2012 - 08:51:12 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:31 UTC