CentOS - Jun 2007 - Strange slowdown occasionally in CentOS 5.0

For some reason, parts of my system are now slowing down to an
agonizing crawl from time to time.  At first it was my VMWare Server
1.04, which became such a pain that I rebooted.  Then, it looked like
it was a network-induced thing because the prime culprit was
SeaMonkey, particularly with multiple tabs open.  I did an ifdown and
ifup on my NIC and that seemed to help for a while.  But then it hit
OO and I'm baffled.  It seems that the second time I open any
document, particularly with OO Writer, the load and save times are
dramatically extended - we're talking five to ten minutes.

I don't even know what information I should include here that might
help, but here's what I have running at the moment:

OS/Hardware Config:
OS : CentOS 5.0 (2.6.16-8.1.6 Linux kernel) x86_64 with latest updates
CPU: AMD Athlon 64 x2 4200+
M/B: ECS NFORCE4M-A
ram: 2GB OCZ DDR2 800 (PC6400)
hds: 160MB + 120MB Maxtor PATA UDMA-133 (160 is the boot drive)
sds: 300GB Seagate SATA-150/300 (/boot and / are here)
        320Gb WD SATA-150/300
vid: geFORCE7100gs (nVidia) PCI-E x16
lpr: Minolta QMS PagePro 1100 laser printer on /dev/lp0
mod: Agere Systems 56k WinModem (Lucent)
nic: onboard (mobo-nvidia)

Any suggestions or more information needed?

Thanks.

mhr

Mark Hull-Richter wrote:
> For some reason, parts of my system are now slowing down to an
> agonizing crawl from time to time.  At first it was my VMWare Server
> 1.04, which became such a pain that I rebooted.  Then, it looked like
> it was a network-induced thing because the prime culprit was
> SeaMonkey, particularly with multiple tabs open.  I did an ifdown and
> ifup on my NIC and that seemed to help for a while.  But then it hit
> OO and I'm baffled.  It seems that the second time I open any
> document, particularly with OO Writer, the load and save times are
> dramatically extended - we're talking five to ten minutes.
I don't think you are imagining things, as I tend to see similar things.

Every time I see it though, it is firefox/seamonkey related ... and
stopping and then restarting the browser has fixed the problem for me.

I also notice that this happens when I have too many tabs open (I
sometimes get as many as 30 tabs open).

I never really looked to see how much memory was in use or the load ...
as I just chalked it up to too many tabs and some kind of memory leak or
another.  As I stated before, closing then reopening the browser does
usually fix it for me.

Since I normally use either firefox 2.0.x or Gran Paradiso
(firefox-3.0alpha) as my browser, I just assumed it was a problem with
that.  (and my issues may be unrelated to yours and caused by my use of
unauthorized browsers)
> I don't even know what information I should include here that might
> help, but here's what I have running at the moment:
> 
> OS/Hardware Config:
> OS : CentOS 5.0 (2.6.16-8.1.6 Linux kernel) x86_64 with latest updates
> CPU: AMD Athlon 64 x2 4200+
> M/B: ECS NFORCE4M-A
> ram: 2GB OCZ DDR2 800 (PC6400)
> hds: 160MB + 120MB Maxtor PATA UDMA-133 (160 is the boot drive)
> sds: 300GB Seagate SATA-150/300 (/boot and / are here)
>        320Gb WD SATA-150/300
> vid: geFORCE7100gs (nVidia) PCI-E x16
> lpr: Minolta QMS PagePro 1100 laser printer on /dev/lp0
> mod: Agere Systems 56k WinModem (Lucent)
> nic: onboard (mobo-nvidia)
> 
> Any suggestions or more information needed?
My hardware is a dual processor Xeon 2.0 ghz, 3gb RAM machine that is my
main workstation.  I will actually take a look at my memory and see if I
can see anything weird the next time it happens.

Thanks,
Johnny Hughes



-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 252 bytes
Desc: OpenPGP digital signature
URL:
<http://lists.centos.org/pipermail/centos/attachments/20070630/2ac4c790/attachment.sig>

On 6/30/07, Johnny Hughes <johnny at centos.org>
wrote:
> Mark Hull-Richter wrote:
> > For some reason, parts of my system are now slowing down to an
> > agonizing crawl from time to time.
>
> I don't think you are imagining things, as I tend to see similar
things.
>
> Every time I see it though, it is firefox/seamonkey related ... and
> stopping and then restarting the browser has fixed the problem for me.
I haven't seen this kind of thing regularly, but the other day I was
trying to do some work that involved converting a couple of relatively
large (3 columns, 3500 rows) HTML tables into a spreadsheet.  I tried
to do this by loading each HTML document into FF, doing a "select all"
and then pasting into oocalc.

This eventually did work, but when I tried to "select all" in FF it
became unresponsive (watched in "top", it was consuming 95%+ of the
CPU) for minutes at a time.  The fist time this happened I was then
able to paste into oocalc, but on the second table I had to try three
times -- it would *seem* to have selected all the rows (as in, the
browser finally became responsive again and showed the whole page
highlighted) but the clipboard was empty (I could not paste).  If it
hadn't worked for the first table I wouldn't have tried as often as I
did on the second one.

I'm tempted to blame this on X (more likely the libraries, but
possibly something to do with the clipboard protocol) rather than on
Firefox.  However, I'm probably biased against Xorg because it still
goes AWOL for long periods at boot time if no network is available
(just this morning I resorted to installing caching-nameserver in
hopes that might help).