Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 01 Jun 2009 10:13:09 +0200
From:      "O. Hartmann" <ohartman@mail.zedat.fu-berlin.de>
To:        Attilio Rao <attilio@freebsd.org>
Cc:        freebsd-current@freebsd.org
Subject:   Re: signifanctly slowdown of FreeBSD 8.0-CURRENT/amd64
Message-ID:  <4A238D95.9020209@mail.zedat.fu-berlin.de>
In-Reply-To: <3bbf2fe10905300951o3beea201j72b6e6b96f85c05a@mail.gmail.com>
References:  <4A2120D5.50300@mail.zedat.fu-berlin.de>	 <3bbf2fe10905300848s6342a7b1l32340baee8e7e8f1@mail.gmail.com>	 <4A215D7F.7020807@mail.zedat.fu-berlin.de>	 <3bbf2fe10905300944r96ae931re4475c78d9be3aa8@mail.gmail.com> <3bbf2fe10905300951o3beea201j72b6e6b96f85c05a@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Attilio Rao wrote:
> 2009/5/30 Attilio Rao <attilio@freebsd.org>:
>   
>> 2009/5/30 O. Hartmann <ohartman@mail.zedat.fu-berlin.de>:
>>     
>>> Attilio Rao wrote:
>>>       
>>>> 2009/5/30 O. Hartmann <ohartman@mail.zedat.fu-berlin.de>:
>>>>
>>>>         
>>>>> Hello.
>>>>> I realized a significant slowdown of FreeBSD 8.0-CURRENT/amd64 on every
>>>>> box I run. I have the most recent FreeBSD 8.0-CURRENT/amd64 with custom
>>>>> kernel and switched off every debugging. I see a drastic slowdown
>>>>> whenever heavy I/O on UFS2 and ZFS partitions is performed and whenever
>>>>> some compilation is done (compiling world and kernel). This is horrible
>>>>> on a single core Athlon64 CPU with 2GB RAM as well as on a 4 core Q6600
>>>>> with 8GB and a server with 2 x 4-cores and 16GB RAM.
>>>>>
>>>>> I can not say clearly whether I/O is the bottleneck. Maybe something
>>>>> with the memory subsystem, when it comes to compiler runs, when no disk
>>>>> I/O is done but the box is still horrible slow. This behaviour occured
>>>>> several weeks ago, not being able to specify it more precisely.
>>>>>
>>>>> AQre there any issues at the moment?
>>>>>
>>>>>           
>>>> Your kernel is compiled from which date?
>>>>
>>>> Thanks,
>>>> Attilio
>>>>
>>>>
>>>>
>>>>         
>>> Most recent, say: yesterday! As well as world.
>>>       
>> Can you try to revert only r193011 and see if something changes?
>>     
>
> Also, did you compile the single-core athlon64 without the option SMP?
>
> Thanks,
> Attilio
>
>
>   

Yes, I did of course compile kernel on UP without option SMP and of
those with more than on logical core with SMP on.

I will try to revert to r193011.

Oliver



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4A238D95.9020209>