Re: Setting up md-raid5: observations, errors, questions

[Bill Davidsen <davidsen@tmr.com>]

Christian Pernegger wrote:
>>  You could try to change cables and such but you've already cc'd linux-ide,
>>  AFAIK it can/could be a chipset/related issue and the guys who work on NCQ
>>  etc are working on the problem is the last I heard..
>>
>>  You should try and narrow down the problem, is it always the same drive
>>  that has that problem every time?  Does it occur if you do a check on the
>>  RAID 5 array or only when building?  etc..
>>     
>
> Assuming you're talking about the HSM violation error ... I got that
> exactly once (yet) in the situation described (sometime after the
> writing phase of badblocks). Nothing since and certainly not the spew
> others are reporting.
>
>   
Are you by any chance running quota?

> My primary concern now is the second problem - seems like any access
> to the raid array makes the machine unusable. I can type/edit a
> command at the bash prompt normally but as soon as I hit enter it just
> hangs there. If the command is trivial, say cat, I might get a result
> a few minutes later.
>
> bonnie++ results are in:
>
> Version 1.03c       ------Sequential Output------ --Sequential Input- --Random-
>                     -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
> Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
> 1024k            4G           16282   5 13359   3           57373   7 149.0   0
>                     ------Sequential Create------ --------Random Create--------
>                     -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
>               files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP
>                  16 +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++
>
> Something is definitely not right ... :-(
>
> As in, an old array of 300GB IDE Maxtors has 4x the seq writes, bus
> capped (105MB/s, plain PCI) reads and 3x the IOs. And it doesn't block
> the machine. Granted, there's the crypto but on my other (non-raid)
> boxes the performance impact just isn't there.
>
> Any help appreciated, as it is the box has expensive-paperweight status.
>
> C.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-raid" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
>   


-- 
Bill Davidsen <davidsen@tmr.com>
  "Woe unto the statesman who makes war without a reason that will still
  be valid when the war is over..." Otto von Bismark