Date: Tue, 12 Oct 1999 11:55:56 +0100 (WEST) From: Joao Pagaime <jpsp@rccn.net> To: "TFC WLAN 97/98 - IST - ext.2269 (8418269)" <wlan@feldspato.ist.utl.pt> Cc: freebsd-questions@FreeBSD.ORG, pagaime@rccn.net Subject: Re: Slow FreeBSD 3.2 Message-ID: <Pine.BSF.4.10.9910121151460.36741-100000@atlas.rccn.net> In-Reply-To: <Pine.LNX.4.05.9910111912040.4978-100000@feldspato.ist.utl.pt>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi
Sorry it's not an answer but rather a new question:
Is there any way to improve (I mean increase) concurrency to
disk accesses ?
Maybe at the file system or kernel levels ?
On Mon, 11 Oct 1999, TFC WLAN 97/98 - IST - ext.2269 (8418269) wrote:
>
> Hello all
>
> We have a Dell 4300 with 2 9GB SCSI disks, 512MB and
> a Pentium II 450 MHz CPU.
>
> But, when ever 'top' reports a process in 'newbuf', 'getblk'
> or 'biowr', the system becomes very slow, sometimes for a
> few seconds. We can't even do a 'ls' properly...
>
> Unfortunately the system ends up on that state whenever
> 'pine' or 'popper' runs on a large mailbox. Of course
> we can't convince people to use 'maildir'.
>
> Since the machine is almost idle - CPU and memory - why does
> this happen ?
>
> Is it a driver/disk problem ?
>
> What do those states reported by top ('newbuf', 'getblk', 'biowr')
> mean ?
>
> Is there some kernel optimization we can do ? Kernel upgrade ?
>
> Thanks,
> Joao
>
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.10.9910121151460.36741-100000>