Re: RH9 Poor performance, stability

From: Zinn (zinn_at_spam.all.people)
Date: 06/12/04 

Date: Fri, 11 Jun 2004 16:12:02 -0600

On Thu, 10 Jun 2004 18:40:55 +0000, Larry I Smith wrote:

> Zinn wrote:
>> I'm getting some rather poor performance out of a Red Hat 9
>> system with PIII 866 Coppermine, and 384 MB PC133 ram. At least
>> part of the problem seems to be memory related. First off,
>> the applications I'm using are sucking up enormous amounts
>> of ram. All the apps I typically keep open are:
>> Mozilla
>> XEmacs
>> Pan
>> Evolution
>> Gnome-terminal
>> Nautilus (kept open by Gnome)
>>
>> However, each of these is sucking up massive amounts of ram, on
>> the order of **dozens of megs** each and I'm typically close to
>> exhausting my physical ram. Mozilla alone can often take up 50 - 100
>> megs. In addition to the above major apps, Gnome opens lots of little
>> things (Gnome-panel, ...) which can take up quite sizeable amounts of
>> ram (like 5 - 15 megs each).
>>
>> Performance is weak, whether it's the memory or something else.
>> Switching between processes is slow, even if physical memory is not
>> exhausted. Killing some apps improves things a little but doesn't always
>> fully restore the system to decent performance.
>>
>> Further, I have stability problems. My system freezes up about once or
>> twice a week or so, probably due to memory exhaustion and thrashing.
>>
>> By comparison, I have Windows 2000 running on a PII 450 with 512 MB ram.
>> Does 128 MB ram make that much difference? It never crashes and it runs
>> **faster** on the whole than Red Hat 9 on the PIII 866. Windows 2000
>> seems to handle memory exhaustion better than Red Hat. For instance,
>> there's a buggy utility program with a memory leak for my wireless card
>> on my Windows 2000 box. After runing for a day or so, the program
>> consumes nearly all system memory. This doesn't phase Win 2000. The
>> system slows down. I get a dialog notifying me that the system is
>> increasing virtual memory. If I kill the program, a few seconds later,
>> everything's back to normal. Why is it so much cleaner than under Linux?
>> The hype I've heard is the opposite.
>>
>> I don't consider these basic activities I'm doing under Linux anything
>> that should be particularly demanding, and I've never seen this kind of
>> massive resource consumption for such ordinary tasks on other operating
>> systems. Is this normal? Is there some system tweek that might ameliorate
>> the problem? Should I upgrade to Fedora?
>>
>> PS. If you think this is a troll, that's fine. Just deposit your
>> comments in /dev/null. On the other hand, if you have something useful
>> to say, I'd be interested in any advice.
>>
>>
>>
>>
>
> Hmm, you haven't provided enough info for anyone
> to give specific recommendations.
>
> By default Linux will cache a lot of data in RAM
> to improve performance. When a particular app
> needs more RAM Linux will reduce the amount of
> RAM used for cache and give the app more RAM.
> The idea is to maximize RAM usage (yes, this
> is very different from Windows).
>
> Unless you have a 'bad' RAM chip, memory
> should not be a problem.
>
> What type of disks (SCSI, IDE, USB, etc)?
> What is the file system (reiser, etc)?
> How are the disks partitioned?
> What are the partition sizes?
> How much swap space?

For primary disk drive, I have a 7200 rpm ATA-100 hard drive with 8.5ms
average seek. There's a secondary drive, but it's on the other IDE
interface.

File system is ext3.

There are 12 partitions on the primary drive:

   Device Boot Start End Blocks Id System
/dev/hda1 * 1 11 88326 6 FAT16
/dev/hda2 12 12 8032+ a OS/2 Boot Manager
/dev/hda3 13 204 1542240 17 Hidden HPFS/NTFS
/dev/hda4 205 3608 27342630 f Win95 Ext'd (LBA)
/dev/hda5 * 205 249 361431 7 HPFS/NTFS
/dev/hda6 * 250 696 3590496 7 HPFS/NTFS
/dev/hda7 * 697 990 2361523+ 83 Linux
/dev/hda8 * 991 1056 530113+ 82 Linux swap
/dev/hda9 * 1057 1560 4048348+ 83 Linux
/dev/hda10 * 1561 2207 5196996 83 Linux
/dev/hda11 2208 2654 3590496 83 Linux
/dev/hda12 3063 3608 4385713+ 83 Linux

My root is mounted on hda11. My /usr is mounted on hda12. My swap is on
hda8, with 530 MB. The other linux partitions are from a previous rh7
installation (whose performance was not *as bad* if I recall). In fact,
here's the output of mount:

/dev/hda11 on / type ext3 (rw)
none on /proc type proc (rw)
none on /dev/pts type devpts (rw,gid=5,mode=620)
none on /dev/shm type tmpfs (rw)
/dev/hda12 on /usr type ext3 (rw)
/dev/hda10 on /mnt/hda10 type ext2 (rw)
/dev/hda7 on /mnt/rh7root type ext2 (rw)
/dev/hda9 on /mnt/rh7usr type ext2 (rw)
/dev/hdb5 on /mnt/spec type ext3 (rw)

> What server daemons are running?

As for services:
[root@samantha Hamilton]# service --status-all
anacron dead but subsys locked
apmd (pid 572) is running...
atd (pid 841) is running...
Configured Mount Points:
------------------------
 
Active Mount Points:
--------------------
cannaserver (pid 702) is running...
crond (pid 713) is running...
cupsd (pid 732) is running...
gpm (pid 681) is running...
httpd (pid 721 720 719 718 717 716 715 714 692) is running...
Table: filter
Chain INPUT (policy ACCEPT)
target prot opt source destination
RH-Lokkit-0-50-INPUT all -- anywhere anywhere
 
Chain FORWARD (policy ACCEPT)
target prot opt source destination
RH-Lokkit-0-50-INPUT all -- anywhere anywhere
 
Chain OUTPUT (policy ACCEPT)
target prot opt source destination
 
Chain RH-Lokkit-0-50-INPUT (2 references)
target prot opt source destination
ACCEPT udp -- clock.redhat.com anywhere udp spt:ntp dpt:ntp
ACCEPT udp -- clock.redhat.com anywhere udp spt:ntp dpt:ntp
ACCEPT udp -- ns5.attbi.com anywhere udp spt:domain dpts:1025:65535
ACCEPT udp -- ns2.attbi.com anywhere udp spt:domain dpts:1025:65535
ACCEPT tcp -- anywhere anywhere tcp dpt:http flags:SYN,RST,ACK/SYN
ACCEPT tcp -- anywhere anywhere tcp dpt:ftp flags:SYN,RST,ACK/SYN
ACCEPT tcp -- anywhere anywhere tcp dpt:ssh flags:SYN,RST,ACK/SYN
ACCEPT tcp -- anywhere anywhere tcp dpt:telnet flags:SYN,RST,ACK/SYN
ACCEPT all -- anywhere anywhere
ACCEPT all -- anywhere anywhere
REJECT tcp -- anywhere anywhere tcp dpts:0:1023 flags:SYN,RST,ACK/SYN reject-with icmp-port-unreachable
REJECT tcp -- anywhere anywhere tcp dpt:nfs flags:SYN,RST,ACK/SYN reject-with icmp-port-unreachable
REJECT udp -- anywhere anywhere udp dpts:0:1023 reject-with icmp-port-unreachable
REJECT udp -- anywhere anywhere udp dpt:nfs reject-with icmp-port-unreachable
REJECT tcp -- anywhere anywhere tcp dpts:x11:6009 flags:SYN,RST,ACK/SYN reject-with icmp-port-unreachable
REJECT tcp -- anywhere anywhere tcp dpt:xfs flags:SYN,RST,ACK/SYN reject-with icmp-port-unreachable
irattach is stopped
No status available for this package
lisa is stopped
rndc: connect failed: connection refused
Configured devices:
lo eth0 ppp0
Currently active devices:
lo eth0
rpc.mountd is stopped
nfsd is stopped
rpc.rquotad is stopped
rpc.statd (pid 504) is running...
nscd is stopped
ntpd (pid 642) is running...
portmap (pid 485) is running...
pxe is stopped
The random data source exists
saslauthd is stopped
sendmail (pid 671 662) is running...
smbd (pid 819) is running...
nmbd (pid 823) is running...
snmpd is stopped
snmptrapd is stopped
sshd (pid 609) is running...
syslogd (pid 463) is running...
klogd (pid 467) is running...
vsftpd is stopped
winbindd is stopped
xfs (pid 810) is running...
xinetd (pid 623) is running...
ypbind is stopped
rpc.yppasswdd is stopped
ypserv is stopped
rpc.ypxfrd is stopped

> What does the system log say after a "freeze"?
> etc, etc, etc....

I don't think it says anything relevant. I don't think the system "knows"
it's frozen. It's just wildly swapping in and out of the hard disk. On
one occasion, my Linux box started freezing up (I could tell because the
mouse started moving jerkily). I ran to my Windows box, which had an
ssh window into it open, and typed something that would kill any process
with a name containing "ozilla" (on the hunch that Mozilla was the culprit).
The characters echoed back at about 1 every 15 minutes. I let the
thing run out of curiosity. About 6 hours later, Mozilla was in
fact killed and the system was usable again. (Another triumph for the
Penguin!). The system didn't think anything anomalous had happened.
It just seems that Red Hat 9 isn't very good at handling stressful memory
situations.

> Provide as much relevant information as possible.
>
> Regards,
> Larry

Relevant Pages

  • Re: How to set up a Linux machine that occupies the minimum memory footprint ?
    ... Of course, if you use 8GB of RAM, or more precisely if you ... Are you saying that if I have a 4GB RAM memory, ... Linux will somehow "know" that it doesn't have ... You've said in another message that your app is ...
    (comp.os.linux.setup)
  • Re: How to set up a Linux machine that occupies the minimum memory footprint ?
    ... Of course, if you use 8GB of RAM, or more precisely if you ... Are you saying that if I have a 4GB RAM memory, ... Linux will somehow "know" that it doesn't have ... You've said in another message that your app is ...
    (comp.os.linux.misc)
  • Re: OT (slightly) swap limits
    ... influence your decision on how much swap was enough. ... processes due to lack of memory. ... I've run machines with 1Gig or more RAM with NO SWAP. ... Linux possible when before on Unix it wasn't? ...
    (Debian-User)
  • Re: oom-killer vs system not reachable
    ... Do you think that Linux might stop to respond or get stuck ... Another situation is that several processes are competing for memory, ... in ram at once. ... is to monitor or log what happens when the system dies like this, ...
    (Debian-User)
  • Re: advice for my new Linux workstation
    ... supports up to 16GB or ram be build today and would latest Debian Linux ... I would expect so (but pick any two from cheap, ... recommend the 64 bit kernel, though, as it manages that memory much ...
    (comp.os.linux.hardware)