Plesk Interface Hangs

[singex]

I recently installed CentOS 5 and plesk 8.2 on this server. It seems to work fine for the most part. But it sometimes hangs and is unresponsive when performing certain functions such as removing domains, migrating accounts, restoring backups, adding ip addresses. It doesn't always become unresponsive when doing these things. But when it does, it just sits at the "Please wait, Loading...". I've tried letting it sit there for an hour or so with no luck. I've also tried closing the browser window but after I submit the login and password it hangs at the login screen. I tried restarting plesk through the shell and it hangs there as well. Rebooting fixes it but it's really starting to bug me. The websites and emails are working fine though.

 

[singex]

UPDATE

This seems odd to me but if I try to log into the plesk interface while I have an active ssh session open it will disconnect my ssh session. I can recreate this every time. I tried disabling both iptables and selinux and it still happens.

 

[singex]

FIXED!

The cause was a cpu lockup:

BUG: soft lockup detected on CPU#2!

Call Trace:
 <IRQ>  [<ffffffff800b2ca3>] softlockup_tick+0xdb/0xed
 [<ffffffff80093424>] update_process_times+0x42/0x68
 [<ffffffff80073d99>] smp_local_timer_interrupt+0x23/0x47
 [<ffffffff8007445b>] smp_apic_timer_interrupt+0x41/0x47
 [<ffffffff80054f2e>] mwait_idle+0x0/0x4a
 [<ffffffff8005bcc2>] apic_timer_interrupt+0x66/0x6c
 <EOI>  [<ffffffff80054f64>] mwait_idle+0x36/0x4a
 [<ffffffff80046fb7>] cpu_idle+0x95/0xb8
 [<ffffffff80073bb7>] start_secondary+0x45a/0x469

The fix was adding "noapic" and "nolapic" to the kernel parameters.

 

[KrazyBob]

*exactly* how and where did you add "noapic" and "nolapic" to the kernel parameters? We seem to have a similar problem with our 3Ware RAID controller and I have never modified a grub.conf file.

Thanks.

 

[singex]

We are also using a 3ware controller on that server. I thought it was a cpu issue. I guess it's a 3ware issue.

From the command line open /etc/grub.conf using vi or whatever. Then add "noapic nolapic" (without the quotes) to the end of the kernel parameter. So your grub.conf would go from this:

default=0
timeout=5
splashimage=(hd0,0)/grub/splash.xpm.gz
hiddenmenu
title Red Hat Enterprise Linux ES (2.6.9-22.ELsmp)
        root (hd0,0)
        kernel /vmlinuz-2.6.9-22.ELsmp ro root=/dev/VolGroup00/LogVol00
        initrd /initrd-2.6.9-22.ELsmp.img

To this:

default=0
timeout=5
splashimage=(hd0,0)/grub/splash.xpm.gz
hiddenmenu
title Red Hat Enterprise Linux ES (2.6.9-22.ELsmp)
        root (hd0,0)
        kernel /vmlinuz-2.6.9-22.ELsmp ro root=/dev/VolGroup00/LogVol00 noapic nolapic
        initrd /initrd-2.6.9-22.ELsmp.img

Of course yours will most likely look differently but this should give you the general idea of how to add arguments to the kernel parameter.

 

[KrazyBob]

Thanks for the reply.

All of our servers are configured and built the same. Only one server seems to have an issue.

Here's my grub.conf

#boot=/dev/sda
default=0
timeout=5
splashimage=(hd0,0)/boot/grub/splash.xpm.gz
hiddenmenu
title Virtuozzo (2.6.9-023stab046.2-enterprise)
        root (hd0,0)
        kernel /boot/vmlinuz-2.6.9-023stab046.2-enterprise ro root=LABEL=/ debug 
        initrd /boot/initrd-2.6.9-023stab046.2-enterprise.img

So I assume that I will add noapic nolapic as follows:

#boot=/dev/sda
default=0
timeout=5
splashimage=(hd0,0)/boot/grub/splash.xpm.gz
hiddenmenu
title Virtuozzo (2.6.9-023stab046.2-enterprise)
        root (hd0,0)
        kernel /boot/vmlinuz-2.6.9-023stab046.2-enterprise ro [color=red]noapic nolapic[/color] root=LABEL=/ debug 
        initrd /boot/initrd-2.6.9-023stab046.2-enterprise.img

although 3Ware only instructed me to add the noapic parameter.

For reference, we are using Super Micro motherboards, 4GB of quality RAM, 3Ware 8006-2 RAID controllers and Seagate NL750 SATA 2 enterprise drives. Only one machine out of 12 has this issue. Off.

 

[singex]

You got it. That should work.

Also, with grub you can have two images specified and change the default image for the next boot only. Here's how you do it:

List the new one first, set "default=1" and add "panic=10" to the kernel parameter. Here's an example of what your grub.conf would look like:

#boot=/dev/sda
default=1
timeout=5
splashimage=(hd0,0)/boot/grub/splash.xpm.gz
hiddenmenu
title Virtuozzo (2.6.9-023stab046.2-enterprise)
        root (hd0,0)
        kernel /boot/vmlinuz-2.6.9-023stab046.2-enterprise ro noapic nolapic panic=10 root=LABEL=/ debug 
        initrd /boot/initrd-2.6.9-023stab046.2-enterprise.img
title Virtuozzo (2.6.9-023stab046.2-enterprise)
        root (hd0,0)
        kernel /boot/vmlinuz-2.6.9-023stab046.2-enterprise ro root=LABEL=/ debug 
        initrd /boot/initrd-2.6.9-023stab046.2-enterprise.img

Then type the following:

grub
savedefault --default=0 --once
quit
reboot

This will reboot your system and attempt to boot using the new image and parameters. If it fails then it will automatically reboot in 10 seconds and boot using the old image and parameters.

If it successfully boots the image you're testing then edit your grub.conf and change "default=1" to "default=0". This is especially helpful when installing a new kernel.

 

[KrazyBob]

I'll head down to the co-lo shortly. Extreme fire weather here in So. California slowing me down.

 

[KrazyBob]

OK. If I boot with a rescue CD am I going to be able to edit the grub.conf? Or do I havfe to mount a partition somehow? This is an advanced topic for me