Hello Folks, We have 2 (identical hardware) busy public 2U rack servers, both have several hundred users & peak email (for example) deliveries at 300/5min during the daytime. Both servers run RAID1 over 2 EIDE disk drives. One server is fine - it never has problems - it is running RH7.3 and CPU loads are normal. The other server has constant disk activity & kjournald at CPU load of 50/60/70% most of the time! this machine is running RH7.2 The same processes & loads are on both machines - can anyone tell me why one machine should have so much kjournald activity & the other not? Is it a difference between 7.2 & 7.3 or must it be another factor? How can I find out? Regards, Nico Morrison nico.morrison@micronicos.com ___________________________________________ Micronicos Limited - London, UK. Tel: +44 20 8870 8849 Fax: +44 20 8870 5290 ___________________________________________
Andrew Morton
2003-Feb-04 02:19 UTC
Re: On very different journalling activity on 2 servers.
Nico Morrison <nico.morrison@micronicos.com> wrote:> > Hello Folks, > > We have 2 (identical hardware) busy public 2U rack servers, both have > several hundred users & peak email (for example) deliveries at 300/5min > during the daytime. > > Both servers run RAID1 over 2 EIDE disk drives. > > One server is fine - it never has problems - it is running RH7.3 and CPU > loads are normal. > > The other server has constant disk activity & kjournald at CPU load of > 50/60/70% most of the time! this machine is running RH7.2 > > The same processes & loads are on both machines - can anyone tell me why one > machine should have so much kjournald activity & the other not? > > Is it a difference between 7.2 & 7.3 or must it be another factor?Usually, this is due to IDE drives running in PIO mode and not DMA mode.> How can I find out? >Look at the output of `hdparm -i': # hdparm -i /dev/hda /dev/hda: Model=Maxtor 5T040H4, FwRev=TAH71DP0, SerialNo=T4J0N9BC Config={ Fixed } RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=57 BuffType=DualPortCache, BuffSize=2048kB, MaxMultSect=16, MultSect=16 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=80043264 IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120} PIO modes: pio0 pio1 pio2 pio3 pio4 DMA modes: mdma0 mdma1 mdma2 UDMA modes: udma0 udma1 *udma2 udma3 udma4 udma5 AdvancedPM=yes: disabled (255) WriteCache=enabled Drive conforms to: ATA/ATAPI-6 T13 1410D revision 0: 1 2 3 4 5 6 See the "*" next to UDMA2? This disk is doing DMA. You can manually force DMA mode, which is considered risky. Or upgrade the kernel, perhaps.
Nico Morrison
2003-Feb-04 08:30 UTC
RE: On very different journalling activity on 2 servers.
Hello Andrew,
Thank you, but unfortunately:
[root@ns5 nico]# /sbin/hdparm -i /dev/hda
/dev/hda:
Model=MAXTOR 6L040J2, FwRev=A93.0500, SerialNo=662240310103
Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs }
RawCHS=16383/16/63, TrkSize=32256, SectSize=21298, ECCbytes=4
BuffType=DualPortCache, BuffSize=1819kB, MaxMultSect=16, MultSect=16
CurCHS=16383/16/63, CurSects=-66060037, LBA=yes, LBAsects=78177792
IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120}
PIO modes: pio0 pio1 pio2 pio3 pio4
DMA modes: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 *udma5 udma6
AdvancedPM=no
Drive Supports : ATA/ATAPI-5 T13 1321D revision 1 : ATA-1 ATA-2 ATA-3 ATA-4
ATA-5
Any other ideas?
Regards,
Nico Morrison
nico.morrison@micronicos.com
___________________________________________
Micronicos Limited - London, UK.
Tel: +44 20 8870 8849 Fax: +44 20 8870 5290
___________________________________________
From: Andrew Morton [mailto:akpm@digeo.com]
Sent: 04 February 2003 02:20
To: Nico Morrison
Cc: ext3-users@redhat.com; nico.morrison@micronicos.com
Subject: Re: On very different journalling activity on 2 servers.
Nico Morrison <nico.morrison@micronicos.com>
wrote:>
> Hello Folks,
>
> We have 2 (identical hardware) busy public 2U rack servers, both have
> several hundred users & peak email (for example) deliveries at 300/5min
> during the daytime.
>
> Both servers run RAID1 over 2 EIDE disk drives.
>
> One server is fine - it never has problems - it is running RH7.3 and CPU
> loads are normal.
>
> The other server has constant disk activity & kjournald at CPU load of
> 50/60/70% most of the time! this machine is running RH7.2
>
> The same processes & loads are on both machines - can anyone tell me
why
one> machine should have so much kjournald activity & the other not?
>
> Is it a difference between 7.2 & 7.3 or must it be another factor?
Usually, this is due to IDE drives running in PIO mode and not DMA mode.
> How can I find out?
>
Look at the output of `hdparm -i':
# hdparm -i /dev/hda
/dev/hda:
Model=Maxtor 5T040H4, FwRev=TAH71DP0, SerialNo=T4J0N9BC
Config={ Fixed }
RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=57
BuffType=DualPortCache, BuffSize=2048kB, MaxMultSect=16, MultSect=16
CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=80043264
IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120}
PIO modes: pio0 pio1 pio2 pio3 pio4
DMA modes: mdma0 mdma1 mdma2
UDMA modes: udma0 udma1 *udma2 udma3 udma4 udma5
AdvancedPM=yes: disabled (255) WriteCache=enabled
Drive conforms to: ATA/ATAPI-6 T13 1410D revision 0: 1 2 3 4 5 6
See the "*" next to UDMA2? This disk is doing DMA.
You can manually force DMA mode, which is considered risky. Or upgrade the
kernel, perhaps.
Nico Morrison
2003-Feb-05 08:29 UTC
RE: On very different journalling activity on 2 servers.
Hello Good People, As we are already using the drives in udma mode, I wondered if either of the settings mentioned here might help & what the implications may be: http://www.redhat.com/support/wpapers/redhat/ext3/tuning.html The server handles a lot of email & several hundred web sites too. Would "data=writeback option" help or not (to reduce journalling activities)? Anything else we can do? Thanks in advance. Regards, Nico Morrison nico.morrison@micronicos.com ___________________________________________ Micronicos Limited - London, UK. Tel: +44 20 8870 8849 Fax: +44 20 8870 5290 ___________________________________________ From: Andrew Morton [mailto:akpm@digeo.com] Sent: 04 February 2003 02:20 To: Nico Morrison Cc: ext3-users@redhat.com; nico.morrison@micronicos.com Subject: Re: On very different journalling activity on 2 servers. Nico Morrison <nico.morrison@micronicos.com> wrote:> > Hello Folks, > > We have 2 (identical hardware) busy public 2U rack servers, both have > several hundred users & peak email (for example) deliveries at 300/5min > during the daytime. > > Both servers run RAID1 over 2 EIDE disk drives. > > One server is fine - it never has problems - it is running RH7.3 and CPU > loads are normal. > > The other server has constant disk activity & kjournald at CPU load of > 50/60/70% most of the time! this machine is running RH7.2 > > The same processes & loads are on both machines - can anyone tell me whyone> machine should have so much kjournald activity & the other not? > > Is it a difference between 7.2 & 7.3 or must it be another factor?Usually, this is due to IDE drives running in PIO mode and not DMA mode.> How can I find out? >Look at the output of `hdparm -i': # hdparm -i /dev/hda /dev/hda: Model=Maxtor 5T040H4, FwRev=TAH71DP0, SerialNo=T4J0N9BC Config={ Fixed } RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=57 BuffType=DualPortCache, BuffSize=2048kB, MaxMultSect=16, MultSect=16 CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=80043264 IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120} PIO modes: pio0 pio1 pio2 pio3 pio4 DMA modes: mdma0 mdma1 mdma2 UDMA modes: udma0 udma1 *udma2 udma3 udma4 udma5 AdvancedPM=yes: disabled (255) WriteCache=enabled Drive conforms to: ATA/ATAPI-6 T13 1410D revision 0: 1 2 3 4 5 6 See the "*" next to UDMA2? This disk is doing DMA. You can manually force DMA mode, which is considered risky. Or upgrade the kernel, perhaps.
Nico Morrison
2003-Feb-05 08:45 UTC
RE: On very different journalling activity on 2 servers.
Hello Andrew,
Upi may be right Sir!
Here:
[root@ns5 nico]# /sbin/hdparm -i /dev/hda
/dev/hda:
Model=MAXTOR 6L040J2, FwRev=A93.0500, SerialNo=662240310103
Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs }
RawCHS=16383/16/63, TrkSize=32256, SectSize=21298, ECCbytes=4
BuffType=DualPortCache, BuffSize=1819kB, MaxMultSect=16, MultSect=16
CurCHS=16383/16/63, CurSects=-66060037, LBA=yes, LBAsects=78177792
IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120}
PIO modes: pio0 pio1 pio2 pio3 pio4
DMA modes: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 *udma5 udma6
AdvancedPM=no
Drive Supports : ATA/ATAPI-5 T13 1321D revision 1 : ATA-1 ATA-2 ATA-3 ATA-4
ATA-5
[root@ns5 nico]# /sbin/hdparm /dev/hda
/dev/hda:
multcount = 16 (on)
I/O support = 1 (32-bit)
unmaskirq = 1 (on)
using_dma = 0 (off)
keepsettings = 0 (off)
nowerr = 0 (off)
readonly = 0 (off)
readahead = 8 (on)
geometry = 4866/255/63, sectors = 78177792, start = 0
[root@ns5 nico]#
Does this mean that dma is in fact off?
The other server (the one with no huge journalling % CPU activity has:
/dev/hda:
multcount = 16 (on)
I/O support = 1 (32-bit)
unmaskirq = 1 (on)
using_dma = 1 (on)
keepsettings = 0 (off)
nowerr = 0 (off)
readonly = 0 (off)
readahead = 8 (on)
geometry = 4866/255/63, sectors = 78177792, start = 0
busstate = 1 (on)
[root@ns3 nico]#
So on that server is it saying dma is on?
Looks like is ON on the server that has no problems and OFF on the one with
loadsa journalling?
What do we need to do & is it safe to turn DMA ON on a busy working public
internet server?
Regards,
Nico Morrison
nico.morrison@micronicos.com
___________________________________________
Micronicos Limited - London, UK.
Tel: +44 20 8870 8849 Fax: +44 20 8870 5290
___________________________________________
-----Original Message-----
From: Andrew Morton [mailto:akpm@digeo.com]
Sent: 05 February 2003 08:36
To: Nico Morrison
Cc: ext3-users@redhat.com
Subject: Re: On very different journalling activity on 2 servers.
Nico Morrison <nico.morrison@micronicos.com>
wrote:>
> Hello Good People,
>
> As we are already using the drives in udma mode
Has that been established?
I haven't seen the output of `hdparm /dev/hda' for both your drives yet.
(That's hdparm _without_ the -i flag).
Nico Morrison
2003-Feb-05 09:03 UTC
RE: On very different journalling activity on 2 servers.
Hello Andrew, Thanks again .... So to summarise: 1. Can take the server down (say on a Sat morning) and try turning on DMS access - but you say this is pretty unsafe. 2. Can upgrade the kernel to match the other machine. Current: Good server (7.3) (DMA yes): kernel-2.4.18-18.7.x Bad server (7.2) (DMA no): kernel-2.4.18-18.7.x So they both use currently the same kernel .... There IS an upgrade available but since they both currently use the same kernel - will upgrading help? What would you recommend? Regards, Nico Morrison nico.morrison@micronicos.com ___________________________________________ Micronicos Limited - London, UK. Tel: +44 20 8870 8849 Fax: +44 20 8870 5290 ___________________________________________ -----Original Message----- From: Andrew Morton [mailto:akpm@digeo.com] Sent: 05 February 2003 08:53 To: Nico Morrison Cc: ext3-users@redhat.com Subject: Re: On very different journalling activity on 2 servers. Nico Morrison <nico.morrison@micronicos.com> wrote:> > Does this mean that dma is in fact off?Yes.> So on that server is it saying dma is on?Yes.> What do we need to do & is it safe to turn DMA ON on a busy working public > internet server?`hdparm -d 1' will force DMA on. No, it is not safe to do on a busy server. It is said to be quite unsafe. If you can schedule some downtime to try that and to test it carefully then maybe. But upgrading the kernel to match the well-behaving machine sounds wiser.
Andrew Morton
2003-Feb-05 09:09 UTC
Re: On very different journalling activity on 2 servers.
Nico Morrison <nico.morrison@micronicos.com> wrote:> > Current: > Good server (7.3) (DMA yes): kernel-2.4.18-18.7.x > Bad server (7.2) (DMA no): kernel-2.4.18-18.7.x > > So they both use currently the same kernel .... >Oh dear. Then the hardware must be different. Or possibly there is a problem with the IDE cabling. If you're sure the harware is identical then try swapping out the IDE cabling. Then try upgrading the kernel. If that doesn't help, you may need to try forcing DMA.
Nico Morrison
2003-Feb-05 10:17 UTC
RE: On very different journalling activity on 2 servers.
Thank you Stephen, We will do a kernel upgrade as you suggest & then see what is what. We'll do it early tomorrow morning & will let you most helpful people know what happened. Thanks again - great suppport. Regards, Nico Morrison nico.morrison@micronicos.com ___________________________________________ Micronicos Limited - London, UK. Tel: +44 20 8870 8849 Fax: +44 20 8870 5290 ___________________________________________ From: Stephen C. Tweedie [mailto:sct@redhat.com] Sent: 05 February 2003 10:13 To: Nico Morrison Cc: 'Andrew Morton'; ext3 users list Subject: RE: On very different journalling activity on 2 servers. Hi, On Wed, 2003-02-05 at 08:45, Nico Morrison wrote:> [root@ns5 nico]# /sbin/hdparm /dev/hda > /dev/hda: > using_dma = 0 (off)> Does this mean that dma is in fact off?Yes.> Looks like is ON on the server that has no problems and OFF on the onewith> loadsa journalling?Yes.> What do we need to do & is it safe to turn DMA ON on a busy working public > internet server?You don't. The trouble is, the kernel usually has a reason for not being in DMA mode. If it can negotiate DMA, it will; but the older of your two systems did not do so, so either that older kernel doesn't know how to drive the IDE controller in DMA mode safely, or it has detected an error on the IDE bus and has backed off to non-DMA mode automatically. In both cases, forcing DMA on manually is not recommended. I'd suggest that you upgrade the 7.2 box to the current errata kernel, which is the same across all the 7.* releases. That should give you the same updated IDE drivers as the 7.3 box has, and will hopefully get you running DMA properly. If it doesn't, then you're going to have to check the logs to see why DMA isn't being negotiated. Cheers, Stephen
Nico Morrison
2003-Feb-06 09:51 UTC
RE: On very different journalling activity on 2 servers.
Hello, I have upgraded the 'bad' machine with the latest 'up2date' kernel, but it is possible that GRUB is still loading the older kernel. Our sysadmin at the dayacentre says the console shows the old kernel kernel-2.4.7-10 This might have happened because the first kernel upgrade I did on this machine I did manually by downloading the .rpm & installing it locally. Now I know to use /use/sbin/up2date instead but ...... How do I tell from the machine (SSH) which kernel it is actually running? Thanks - sorry to be so ignorant. Regards, Nico Morrison nico.morrison@micronicos.com ___________________________________________ Micronicos Limited - London, UK. Tel: +44 20 8870 8849 Fax: +44 20 8870 5290 ___________________________________________ From: Stephen C. Tweedie [mailto:sct@redhat.com] Sent: 05 February 2003 10:13 To: Nico Morrison Cc: 'Andrew Morton'; ext3 users list Subject: RE: On very different journalling activity on 2 servers. Hi, On Wed, 2003-02-05 at 08:45, Nico Morrison wrote:> [root@ns5 nico]# /sbin/hdparm /dev/hda > /dev/hda: > using_dma = 0 (off)> Does this mean that dma is in fact off?Yes.> Looks like is ON on the server that has no problems and OFF on the onewith> loadsa journalling?Yes.> What do we need to do & is it safe to turn DMA ON on a busy working public > internet server?You don't. The trouble is, the kernel usually has a reason for not being in DMA mode. If it can negotiate DMA, it will; but the older of your two systems did not do so, so either that older kernel doesn't know how to drive the IDE controller in DMA mode safely, or it has detected an error on the IDE bus and has backed off to non-DMA mode automatically. In both cases, forcing DMA on manually is not recommended. I'd suggest that you upgrade the 7.2 box to the current errata kernel, which is the same across all the 7.* releases. That should give you the same updated IDE drivers as the 7.3 box has, and will hopefully get you running DMA properly. If it doesn't, then you're going to have to check the logs to see why DMA isn't being negotiated. Cheers, Stephen
Nico Morrison
2003-Feb-06 10:40 UTC
RE: On very different journalling activity on 2 servers.
Hello All,
Well - I am baffled as our techies wrote:
<SNIP>
BB3 is running the GRUB bootloader.
The kernel updates has updated the grub config file which now shows 3
boot options for kernels:
2.4.18-24.7.x
2.4.18-18.7.x
2.4.7-10
default option is set to the first entry.
However when the server boots the only option displayed by GRUB is
2.4.7-10
This is confirmed when the server boots and the kernel is displayed as
2.4.7-10
</SNIP>
Any ideas why GRUB does not show all 3 kernels & then boot the latest?
Here is /boot/grub.conf
<snip>
[root@ns5 grub]# cat grub.conf
# grub.conf generated by anaconda
#
# Note that you do not have to rerun grub after making changes to this file
# NOTICE: You do not have a /boot partition. This means that
# all kernel and initrd paths are relative to /, eg.
# root (hd0,1)
# kernel /boot/vmlinuz-version ro root=/dev/md0
# initrd /boot/initrd-version.img
#boot=/dev/hda
default=0
timeout=10
splashimage=(hd0,1)/boot/grub/splash.xpm.gz
title Red Hat Linux (2.4.18-24.7.x)
root (hd0,1)
kernel /boot/vmlinuz-2.4.18-24.7.x ro root=/dev/md0
initrd /boot/initrd-2.4.18-24.7.x.img
title Red Hat Linux (2.4.18-18.7.x)
root (hd0,1)
kernel /boot/vmlinuz-2.4.18-18.7.x ro root=/dev/md0
initrd /boot/initrd-2.4.18-18.7.x.img
title Red Hat Linux (2.4.7-10)
root (hd0,1)
kernel /boot/vmlinuz-2.4.7-10 ro root=/dev/md0
initrd /boot/initrd-2.4.7-10.img
</snip>
Help - baffled!
Regards,
Nico Morrison
nico.morrison@micronicos.com
___________________________________________
Micronicos Limited - London, UK.
Tel: +44 20 8870 8849 Fax: +44 20 8870 5290
___________________________________________
-----Original Message-----
From: Stephen DeBrass [mailto:debrass@staff.singnet.com.sg]
Sent: 06 February 2003 10:20
To: Nico Morrison
Cc: 'Stephen C. Tweedie'; 'Andrew Morton'; ext3 users list
Subject: RE: On very different journalling activity on 2 servers.
On Thu, 6 Feb 2003, Nico Morrison wrote:
> Hello,
>
> I have upgraded the 'bad' machine with the latest 'up2date'
kernel, but it
> is possible that GRUB is still loading the older kernel. Our sysadmin at
the> dayacentre says the console shows the old kernel kernel-2.4.7-10
>
> This might have happened because the first kernel upgrade I did on this
> machine I did manually by downloading the .rpm & installing it locally.
>
> Now I know to use /use/sbin/up2date instead but ......
You may need to add an entry to /etc/grub.conf that looks something like:
title Red Hat Linux (2.4.19)
root (hd0,0)
kernel /vmlinuz-2.4.19 ro root=/dev/hda7
I believe grub picks defaults to booting the first entry it finds, someone
correct me if I'm wrong, I'm not all too familiar with grub...
> How do I tell from the machine (SSH) which kernel it is actually running?
uname -a
> Thanks - sorry to be so ignorant.
No problem :)
> Regards,
> Nico Morrison
> nico.morrison@micronicos.com
> ___________________________________________
> Micronicos Limited - London, UK.
> Tel: +44 20 8870 8849 Fax: +44 20 8870 5290
> ___________________________________________
>
> From: Stephen C. Tweedie [mailto:sct@redhat.com]
> Sent: 05 February 2003 10:13
> To: Nico Morrison
> Cc: 'Andrew Morton'; ext3 users list
> Subject: RE: On very different journalling activity on 2 servers.
>
>
> Hi,
>
> On Wed, 2003-02-05 at 08:45, Nico Morrison wrote:
>
> > [root@ns5 nico]# /sbin/hdparm /dev/hda
> > /dev/hda:
> > using_dma = 0 (off)
>
> > Does this mean that dma is in fact off?
>
> Yes.
>
> > Looks like is ON on the server that has no problems and OFF on the one
> with
> > loadsa journalling?
>
> Yes.
>
> > What do we need to do & is it safe to turn DMA ON on a busy
working
public> > internet server?
>
> You don't.
>
> The trouble is, the kernel usually has a reason for not being in DMA
> mode. If it can negotiate DMA, it will; but the older of your two
> systems did not do so, so either that older kernel doesn't know how to
> drive the IDE controller in DMA mode safely, or it has detected an error
> on the IDE bus and has backed off to non-DMA mode automatically. In
> both cases, forcing DMA on manually is not recommended.
>
> I'd suggest that you upgrade the 7.2 box to the current errata kernel,
> which is the same across all the 7.* releases. That should give you the
> same updated IDE drivers as the 7.3 box has, and will hopefully get you
> running DMA properly. If it doesn't, then you're going to have to
check
> the logs to see why DMA isn't being negotiated.
>
> Cheers,
> Stephen
>
>
>
> _______________________________________________
> Ext3-users mailing list
> Ext3-users@redhat.com
> https://listman.redhat.com/mailman/listinfo/ext3-users
>
Nico Morrison
2003-Feb-06 10:45 UTC
RE: On very different journalling activity on 2 servers.
Also here is the contents of /boot - I am concerned about this symblink: lrwxrwxrwx 1 root root 19 Feb 6 09:27 System.map -> System.map-2.4.7-10 as all the other symblinks ppint to the latest version: <SNIP> [root@ns5 boot]# ls -al total 10908 drwxr-xr-x 3 root root 4096 Feb 6 09:27 . drwxr-xr-x 23 root root 4096 Feb 6 09:56 .. -rw-r--r-- 1 root root 5824 Jun 25 2001 boot.b -rw-r--r-- 1 root root 612 Jun 25 2001 chain.b -rw-r--r-- 1 root root 42268 Nov 14 01:50 config-2.4.18-18.7.x -rw-r--r-- 1 root root 42257 Jan 31 12:20 config-2.4.18-24.7.x drwxr-xr-x 2 root root 4096 Feb 6 09:16 grub -rw-r--r-- 1 root root 126906 Dec 16 10:34 initrd-2.4.18-18.7.x.img -rw-r--r-- 1 root root 127368 Feb 6 09:15 initrd-2.4.18-24.7.x.img -rw-r--r-- 1 root root 337546 Feb 22 2002 initrd-2.4.7-10.img lrwxrwxrwx 1 root root 14 Dec 16 10:33 kernel.h -> kernel.h-2.4.9 -rw-r--r-- 1 root root 405 Dec 16 10:33 kernel.h-2.4.9 -rw-r--r-- 1 root root 23108 Jun 25 2001 message lrwxrwxrwx 1 root root 25 Feb 6 09:15 module-info -> module-info-2.4.18-24.7.x -rw-r--r-- 1 root root 15436 Nov 14 01:50 module-info-2.4.18-18.7.x -rw-r--r-- 1 root root 15436 Jan 31 12:20 module-info-2.4.18-24.7.x -rw-r--r-- 1 root root 13598 Sep 6 2001 module-info-2.4.7-10 -rw-r--r-- 1 root root 640 Jun 25 2001 os2_d.b lrwxrwxrwx 1 root root 19 Feb 6 09:27 System.map -> System.map-2.4.7-10 -rw-r--r-- 1 root root 490460 Nov 14 01:50 System.map-2.4.18-18.7.x -rw-r--r-- 1 root root 490643 Jan 31 12:20 System.map-2.4.18-24.7.x -rw-r--r-- 1 root root 435039 Sep 6 2001 System.map-2.4.7-10 -rwxr-xr-x 1 root root 2983920 Nov 14 01:50 vmlinux-2.4.18-18.7.x -rwxr-xr-x 1 root root 2986554 Jan 31 12:20 vmlinux-2.4.18-24.7.x lrwxrwxrwx 1 root root 21 Feb 6 09:15 vmlinuz -> vmlinuz-2.4.18-24.7.x -rw-r--r-- 1 root root 1064284 Nov 14 01:50 vmlinuz-2.4.18-18.7.x -rw-r--r-- 1 root root 1064546 Jan 31 12:20 vmlinuz-2.4.18-24.7.x -rw-r--r-- 1 root root 802068 Sep 6 2001 vmlinuz-2.4.7-10 </SNIP> Is there some reason why System.map -> System.map-2.4.7-10 instead of System.map-2.4.18-24.7.x ?? Thank you in advance. Regards, Nico Morrison nico.morrison@micronicos.com ___________________________________________ Micronicos Limited - London, UK. Tel: +44 20 8870 8849 Fax: +44 20 8870 5290 ___________________________________________ From: Stephen DeBrass [mailto:debrass@staff.singnet.com.sg] Sent: 06 February 2003 10:20 To: Nico Morrison Cc: 'Stephen C. Tweedie'; 'Andrew Morton'; ext3 users list Subject: RE: On very different journalling activity on 2 servers. On Thu, 6 Feb 2003, Nico Morrison wrote:> Hello, > > I have upgraded the 'bad' machine with the latest 'up2date' kernel, but it > is possible that GRUB is still loading the older kernel. Our sysadmin atthe> dayacentre says the console shows the old kernel kernel-2.4.7-10 > > This might have happened because the first kernel upgrade I did on this > machine I did manually by downloading the .rpm & installing it locally. > > Now I know to use /use/sbin/up2date instead but ......You may need to add an entry to /etc/grub.conf that looks something like: title Red Hat Linux (2.4.19) root (hd0,0) kernel /vmlinuz-2.4.19 ro root=/dev/hda7 I believe grub picks defaults to booting the first entry it finds, someone correct me if I'm wrong, I'm not all too familiar with grub...> How do I tell from the machine (SSH) which kernel it is actually running?uname -a> Thanks - sorry to be so ignorant.No problem :)> Regards, > Nico Morrison > nico.morrison@micronicos.com > ___________________________________________ > Micronicos Limited - London, UK. > Tel: +44 20 8870 8849 Fax: +44 20 8870 5290 > ___________________________________________ > > From: Stephen C. Tweedie [mailto:sct@redhat.com] > Sent: 05 February 2003 10:13 > To: Nico Morrison > Cc: 'Andrew Morton'; ext3 users list > Subject: RE: On very different journalling activity on 2 servers. > > > Hi, > > On Wed, 2003-02-05 at 08:45, Nico Morrison wrote: > > > [root@ns5 nico]# /sbin/hdparm /dev/hda > > /dev/hda: > > using_dma = 0 (off) > > > Does this mean that dma is in fact off? > > Yes. > > > Looks like is ON on the server that has no problems and OFF on the one > with > > loadsa journalling? > > Yes. > > > What do we need to do & is it safe to turn DMA ON on a busy workingpublic> > internet server? > > You don't. > > The trouble is, the kernel usually has a reason for not being in DMA > mode. If it can negotiate DMA, it will; but the older of your two > systems did not do so, so either that older kernel doesn't know how to > drive the IDE controller in DMA mode safely, or it has detected an error > on the IDE bus and has backed off to non-DMA mode automatically. In > both cases, forcing DMA on manually is not recommended. > > I'd suggest that you upgrade the 7.2 box to the current errata kernel, > which is the same across all the 7.* releases. That should give you the > same updated IDE drivers as the 7.3 box has, and will hopefully get you > running DMA properly. If it doesn't, then you're going to have to check > the logs to see why DMA isn't being negotiated. > > Cheers, > Stephen > > > > _______________________________________________ > Ext3-users mailing list > Ext3-users@redhat.com > https://listman.redhat.com/mailman/listinfo/ext3-users >