Plesk 8.0 causes server crash once every 1-3 days.

[renegadeavenger]

Hey Steve, what kernel were you using before when you were experiencing some crashes?

 

[PE-Steve]

Originally posted by renegadeavenger
Hey Steve, what kernel were you using before when you were experiencing some crashes?

A custom compiled 2.6.9-22. It started off with a nice 70 day uptime with literally 0 load, until the server decided to crash. This wasn't long after upgrading to Plesk 8.0.0.

 

[acer]

Originally posted by PE-Steve
I've been having these problems as well.

I'm running CentOS 4.3 with Plesk 8.0.1. Some of my servers have been randomly crashing and there are no logs of anything going wrong before the crash. While this may be related to a motherboard or powersupply issue that we are looking into, it seems rather odd that others are having the same problem.

The servers reply to PING, although every service on the box is not accessible. Plesk, httpd, mysql, ssh, everything does not work.

Has anyone experience any issues related to the above paragraph (about only PING answering my requests)?

SWSoft, you are REALLY starting to piss people off, especially me, and I've been a HUGE supporter of your product since the day I started using it. Fix these stupid errors or you will have MANY people looking into other control panels. Direct Admin is definitely looking like a better option for me at this point. This is definitely not the only problem I've had. I'm about to vomit over the quality of products that you release.

EDIT: And yes, it never crashed while using Plesk 7.5.4 reloaded.

Regards,
Steve

Same exacty problem here. My server keeps crashing the same way which only just happened today.

The last message log seems to have took place at 7:16 AM. And this is my crontab about 7:17:
17 7 * * * /usr/local/psa/admin/bin/php /usr/local/psa/admin/plib/report/autoreport.php --auto daily >/dev/null 2>&1
17 7 * * 1 /usr/local/psa/admin/bin/php /usr/local/psa/admin/plib/report/autoreport.php --auto weekly >/dev/null 2>&1
17 7 1 * * /usr/local/psa/admin/bin/php /usr/local/psa/admin/plib/report/autoreport.php --auto monthly >/dev/null 2>&1

Could Auto Report have been causing the memory problem and make the services fail ??

 

[renegadeavenger]

What version of the kernel are you using as well??

well based on my findings, Suse 9.3 and Plesk 8.0 is Stable ( at least so far ).

 

[acer]

I'm using Plesk 8.01 with the latest updates.

And my kernel is 2.6.17.3-tss compiled by TSS.

Server crashed again today and afterward I started getting this message:

smartd[1977]: Device: /dev/hda, FAILED SMART self-check. BACK UP DATA NOW!

I backuped up immediately but I think my harddrive is dying.

 

[renegadeavenger]

yeah, try replacing HD and see if you are still crashing.

 

[fishnyc22]

Hey again... So, I updated to 8.0.1 and thought I was in the clear, but I crashed this afternoon.

I started digging around the logs to (the ones I know how to fine in /var/log/ and noticed something that might be odd.

In my cron log, every 15min the backupmng line appears TWO times in a row, for example:

The last thing I see before the crash is this line 2x. And every line like it before and after occurs 2x as well.

Aug 7 17:15:01 u15678380 crond[10248]: (root) CMD (/usr/local/psa/admin/sbin/backupmng >/dev/null 2>&1)
Aug 7 17:15:01 u15678380 crond[10251]: (root) CMD (/usr/local/psa/admin/sbin/backupmng >/dev/null 2>&1)

Earlier in the thread I had asked about what "backupmng" does and someone mentioned that it manages the schedules backups. Since upgrading to 8.0.1 last week (having FTP backups), I have scheduled a backup on one of my domains (I only have 3 domains on the server).

Anyone have any thoughts on what is causing the crash? Any help would be appreciated.

EDIT: I should mention that I just went in an paused the cron job for backupmng and when i did so, a paused (RED X) version appeared at the bottom of the list of crons, BUT here was still an active one at the top. WEIRD!

EDIT2: after turning off the 1 cron job, I now only have that line 1x in my log. Maybe it won't crash now. We'll see.

Fish ><>

 

[fishnyc22]

Just a followup. I tried to delete the paused cron job since it was showing the other one being active. When I did that, both are now gone.

Looking at my logs, I'm seeing this now:

Aug 7 19:46:30 u15678380 crontab[4270]: (root) LIST (root)
Aug 7 19:46:37 u15678380 crontab[4277]: (root) LIST (root)
Aug 7 19:50:01 u15678380 crontab[4363]: (root) BEGIN EDIT (root)
Aug 7 19:51:42 u15678380 crontab[4363]: (root) END EDIT (root)

 

[nblackburn]

same here , I have chage my os and plesk version.

Server : IBM Xseries 360 , 1rx 4 x 1.4ghz xeon cpu
3 go ram
3 x 72 gb scsi raid 5 hdd

OS : Redhat Linux Enterprise 4 AS

Plesk : 8.0.1

It crash once a day
I can ping the server but no screen and nothing work.
The server dosen't have a warning active led.

I will try now to desactive crond and drweb.

If it dosen't work, I will install the same software on another server for see if its ont an hardware problem.

I will let you know....

 

[nblackburn]

Desactivate crond and drweb dosen't nothing.

I will try centos

 

[Seibertron]

Originally posted by acer
Same exacty problem here. My server keeps crashing the same way which only just happened today.

The last message log seems to have took place at 7:16 AM. And this is my crontab about 7:17:
17 7 * * * /usr/local/psa/admin/bin/php /usr/local/psa/admin/plib/report/autoreport.php --auto daily >/dev/null 2>&1
17 7 * * 1 /usr/local/psa/admin/bin/php /usr/local/psa/admin/plib/report/autoreport.php --auto weekly >/dev/null 2>&1
17 7 1 * * /usr/local/psa/admin/bin/php /usr/local/psa/admin/plib/report/autoreport.php --auto monthly >/dev/null 2>&1

Could Auto Report have been causing the memory problem and make the services fail ??

My server is having the same problem on an almost daily basis around 7am. I thought to look at the cron tasks on a whim this morning and noticed that there are 3 cron tasks scheduled to run at 7:17am. I have removed this reporting tools for the time being since I can't think of a time in the past 3 months that I have used the reporting tools from Plesk. I'll let you guys know if this resolved my server's problem.

 

[CruzMark]

A note here for FreeBSD 6.0 users. We were having problems with the Plesk server running 8.0 crashing as well.

After a few weeks, I tracked it down to a bug in the quota.c program in the kernel.

To fix it, you have to apply 3 kernel patches. I've included them here for your use, and I take no responsibility if they hose your server.

You can find the patches at the FreeBSD site as well.
*******************************************************************

Index: ufs_quota.c
===================================================================
RCS file: /home/ncvs/src/sys/ufs/ufs/ufs_quota.c,v
retrieving revision 1.77
retrieving revision 1.80
diff -u -u -r1.77 -r1.80
--- ufs_quota.c 9 Jan 2006 20:42:19 -0000 1.77
+++ ufs_quota.c 22 Feb 2006 06:12:53 -0000 1.80
@@ -429,8 +429,9 @@
quotaoff(td, mp, type);
ump->um_qflags[type] |= QTF_OPENING;
mp->mnt_flag |= MNT_QUOTA;
- ASSERT_VOP_LOCKED(vp, "quotaon");
+ vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, td);
vp->v_vflag |= VV_SYSTEM;
+ VOP_UNLOCK(vp, 0, td);
*vpp = vp;
/*
* Save the credential of the process that turned on quotas.
@@ -535,8 +536,9 @@
}
MNT_IUNLOCK(mp);
dqflush(qvp);
- ASSERT_VOP_LOCKED(qvp, "quotaoff");
+ vn_lock(qvp, LK_EXCLUSIVE | LK_RETRY, td);
qvp->v_vflag &= ~VV_SYSTEM;
+ VOP_UNLOCK(qvp, 0, td);
error = vn_close(qvp, FREAD|FWRITE, td->td_ucred, td);
ump->um_quotas[type] = NULLVP;
crfree(ump->um_cred[type]);
@@ -748,7 +750,7 @@
MNT_ILOCK(mp);
continue;
}
- error = vget(vp, LK_EXCLUSIVE | LK_NOWAIT | LK_INTERLOCK, td);
+ error = vget(vp, LK_EXCLUSIVE | LK_INTERLOCK, td);
if (error) {
MNT_ILOCK(mp);
if (error == ENOENT) {

********************************************

Edit: Link to FreeBSD site for patches -
http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/ufs/ufs/ufs_quota.c

You want 1.77, 1.78, and 1.79, and possibly 1.80

Good luck.

 

[fishnyc22]

Ugh... 1and1 had replaced my hardware (moved my HD to a new machine) to try and solve this problem.

The machine would reboot several times a day AND sometimes crash (about 1 or 2 times very week).

I thought the problem was solved but the machine is STILL rebooting and it just crashed for the first time in about 3 weeks.

Looking in the logs it seems to have crashed at around 8:02am. I only see these 2 lines called around that time.

Jan 6 08:00:01 SERVERNAME crond[22779]: (root) CMD (/usr/local/psa/admin/sbin/backupmng >/dev/null 2>&1)
Jan 6 08:01:01 SERVERNAME crond[22797]: (root) CMD (run-parts /etc/cron.hourly)

Any idea whats happening. will updating to 8.1 (I'm running 8.0.1) help at all?

 

[fishnyc22]

My server still reboots muliple times a day. I'm going nuts. they've replaced the hardware (except the HD) and it still happens.

Recently I've been keeping "top" open in my terminal and watching it. the FREE memory is gradually getting less and less and then the server dies.

Plesk 8.0.1 FC4

If I didnt shave my head I'd have pulled my hair out anyway.

Does anyone know whats up? What could be causing this? Will 8.1 fix anything. I'm afraid to upgrade.

 

[Limedrink]

Hey guys, I think we can help each other.

I have always wondered why my server would go months with uptime, no load problems, and then stop responding.

My datacenter thought it was a power supply failure too. It's not.

My server responds to pings through the IP in the "un-responsive" state, no services are accessible, and no messages are left on the screen. No logs left behind to describe the problem.

It wasn't until the other day that I went to restore an account through the Backup Manager. The moment I clicked 'Restore' my SSH session locked up and the server was INSTANTLY unresponsive.

Disabling the cron makes sure the system doesn't do anymore backups. Users who already have backups in the system and try to restore them will lock up your system.

I was so pissed off the other day, I manually went into the PSA web-file and renamed the backup folder, so that nobody can even get to the page and launch the restore function.

I'm thinking it's related to the kernel. I'm running Fedora Core 1.

Linux 2.6.14 #1 SMP

I could attempt a kernel upgrade, but first I'd like to know if any of you had success with upgrading to PSA 8.1.0 and having this problem resolved.


Hope that helps... We need to find a solution to this.

 

[fishnyc22]

I'm beginning to think its the kernel as well. Please see THIS POST I just threw up here yesterday.

My cousin is a Linux admin (not to familiar with plesk) and has spent the week digging around. He's pretty stumped as well but all thinks are leading to the kernal.

what hosting are you with? 1&1?

Thx

Fish

 

[Limedrink]

Currently I'm with Layered Tech. I switched off of 1&1 over a year ago. They were a big headache.

I'm really curious to see if Plesk 8.1.0 fixes this problem. That would really save a lot of work.

Can anyone confirm?

 

[fishnyc22]

I'd be curious to know if 8.1 solves that problem but I'm hesitant to try it.

Are you happy with Layered Tech? is it virtual server or do you have your own box?

 

[Limedrink]

I'm EXTREMELY happy with Layered Tech. Their support is absolutely top-notch, and their business ethics are amazing. I have my own dedicated box with them -- switched over from 1&1. Reboots are usually handled within 10 min. Techs are VERY friendly, helpful, and most importantly, they know what they are talking about. Their support goes above and beyond "self-managed" and their prices are the best in the market.

1&1 gave me a really hard time with everything and I couldn't stand waiting over 48 business hours for a support email to be responded to (and STILL didn't answer my question). They charge you for bandwidth overages, however, the only method they give you to monitor your usage is through simple IPtables chains that are buggy, get erased on reboots, and plain don't work with IPtables firewalls. The last straw was when I needed to upgrade my box and they told me I'd had to cancel my current server and order one of the newer packages.

They gave me a big headache.

I'll never regret moving over and I don't think you will either.

If you need anymore info, just let me know.

 

[fishnyc22]

thanks for the info. I'll def look into them. It is the worst customer service around. its insane how awful they are.

Let me know if 8.1 solves your restart/crash issue.

Thanks

D