phpmyadmin export problems

[Peter-]

[SOLVED]

My plesk strikes again. I'm trying to export a database that I need and ... bam!

"502 Bad Gateway nginx".

Cleared cookies, tried different browsers... it simply refuses to behave. What do I do now?!

-------------------------------------------------------------------------------------------

[SOLUTION] to save you from reading the whole thread:

Apparently, it's that time of the month when Plesk has a problem with everything.

After abdi's suggestion, I increased my buffers:

#cat /etc/sw-cp-server/config

...

fastcgi_buffers 32 32k; // was 16 16k
fastcgi_buffer_size 64k; // was 32k

...

For a while, plesk admin stopped working. It was because of the timezone. For some dumb reason. It was working before, not working after changing the buffers. I don't see the reason, but ok.

So, next, I had to remove the timezone data.

#yum reinstall tzdata

Next, restart Plesk:
#/etc/init.d/sw-engine restart

Worked.

Waiting for the next server fault.

------------------------------------------------------------------------------------

... and it's back.

I need my timezone to be GMT. If I change it with those buffers set to my second values, I'm getting the session error. Nice job, Plesk!

I'm also still getting the gateway error from time to time.

 

[abdi]

Most times this occurs if the db connections are so slow that your max_execution_time expires (ends) before the db export task execution is complete ...

 

[Peter-]

Hi,

My server is almost empty. Nothing runs slow.

 

[abdi]

Perhaps the DB is huge that again takes that while ....
You need to check with Plesk's Nginx time out (increase it if necessary), PHP max_execution_time ...

You can also look through the logs for the exact cause ...

You could also just dump your db from shell easily with

mysqldump -uadmin -p`cat /etc/psa/.psa.shadow` your_db_name > your_db_name.sql

 

[Peter-]

Not huge either. It's 8MB. This also occurs if I try to export an empty table.

 

[abdi]

Try this fix:

1. Take backup of "/etc/sw-cp-server/config"
2. Increase the following buffer sizes in the file "/etc/sw-cp-server/config".

fastcgibuffers 32 32k;
fastcgibuffer_size 64k;

After editing the file, restart the below services:

/etc/init.d/sw-cp-server restart
/etc/init.d/sw-engine restart

 

[abdi]

Also this was a known bug which was resolved with 11.5.30 MU#14

http://kb.parallels.com/117127
http://forum.parallels.com/showthread.php?289060-Plesk-11-5-502-bad-gateway

 

[Peter-]

Mine is Panel version 11.5.30 Update #13, last updated at Aug 23, 2013 09:52 AM

Last time I updated my plesk I had to delete all my subscriptions... I'm starting to believe it was a bad idea...

 

[abdi]

Last time I updated my plesk I had to delete all my subscriptions... I'm starting to believe it was a bad idea...

Am so sorry to hear that!

Well, give a try to the fix, and if it fails to work then resort to the SHELL dumps else you update to MU#14

 

[Peter-]

I did the changes. Got this error in admin:


ERROR: Session_Exception
Up Level
Session expired
Search for related Knowledge Base articles

I reverted, re-restarted... and still getting it. What do I do now? I cannot access my admin at all

 

[abdi]

tail -f /usr/local/psa/admin/logs/panel.log

For the exact errors ...

 

[Peter-]

tail -f /usr/local/psa/admin/logs/panel.log
[25-Sep-2013 09:53:47 UTC] Session_Exception: Session expired
file: /usr/local/psa/admin/plib/Session.php
line: 410
code: 0
trace: #0 /usr/local/psa/admin/plib/cmd_loginup.php(135): Session->init()
#1 /usr/local/psa/admin/plib/cmd_loginup.php(278): createSessionCommon(1, 'admin', true, '79.112.110.6', NULL, NULL)
#2 /usr/local/psa/admin/plib/cmd_loginup.php(68): createSessionAdmin('admin', '***', false, NULL, true)
#3 /usr/local/psa/admin/htdocs/login_up.php3(45): createSession('admin', '***', false, false)
#4 {main}

 

[abdi]

One of the reasons of such behavior is wrong timezone. Check output of "date" command and correct timezone if needed. Check that /usr/local/psa/admin/conf/php.ini does not contain date.timezone parameter (remove if any).

And if yours is a VPS

Changed timezone on VPS host server back to proper setting using the following.

http://www.wikihow.com/Change-the-Timezone-in-Linux

Rebooted Plesk VPS

 

[Peter-]

[root@server sw-cp-server]# date
Wed Sep 25 10:03:24 GMT 2013

It's a dedicated server. Here's my /usr/local/psa/admin/conf/php.ini:

short_open_tag = On
y2k_compliance = Off
output_buffering = Off
max_execution_time = 600
max_input_time = 600
memory_limit = 128M
max_file_uploads = 99999
max_input_vars = 2000

error_reporting = E_ALL & ~E_NOTICE & ~E_STRICT & ~E_DEPRECATED
log_errors = On
display_errors = Off
warn_plus_overloading = Off
expose_php = Off

variables_order = "ECGPS"
register_argc_argv = On
post_max_size = 2147483647
magic_quotes_gpc = Off
magic_quotes_runtime = Off

include_path = "/usr/local/psa/admin/plib:/usr/local/psa/admin/externals"
upload_tmp_dir = "/tmp"
upload_max_filesize = 2147483647

apc.stat = 1
apc.shm_size = 100M
apc.max_file_size = 500K
apc.slam_defense = 0
apc.enabled = 1

swkey.repository_dir = "/etc/sw/keys"

psasem.semfile = "/usr/local/psa/var/psasem.sem"

zend_extension = "/usr/lib64/php/modules/sw-engine/ioncube_loader_lin_5.4.so"

 

[Peter-]

How come it was working until I changed the buffer size and now it isn't? With the same timezone settings?!

I posted my php.ini content but it says it needs approval. No timezone referrence.

How do I solve this please?

 

[abdi]

There are several reasons that can use that ...(most times requires handson investigations), however please check the following:

1. Ensure the your hard-disk is not full (df -h)
2. Ensure the session folder is writable by plesk ..
3. If you don't loose much, do a reboot of your server ..

 

[Peter-]

What???

It's EMPTY. Everything was working 30 mins ago except that bad gateway error. My SSD is 250GB, hosts about 8 domains, most of them with a couple MBs. A full backup has 80MB of data. CPU shows 0.00 load. Everything was installed at first server boot (and the only server boot til now).

Why would I reboot the server ??

Please, if you know how to restart plesk only or whatever else I need to do, let me know. If you don't stop posting from the manuals

Output for df -h

Filesystem Size Used Avail Use% Mounted on
/dev/md2 206G 3.8G 192G 2% /
tmpfs 16G 0 16G 0% /dev/shm
/dev/md1 496M 31M 440M 7% /boot

I don't know how to check if session folder is writable by plesk, but it was 30 mins ago and I haven't changed anything except that buffer size you told me. Now it's reverted and it doesn't work. Good job, plesk.

---------------------------

~# /etc/init.d/psa stop
~# /etc/init.d/psa start

Didn't do it.

 

[abdi]

stop posting from the manuals

Sure, I have stopped!

 

[Peter-]

Thanks.

Now, can you help me with my problem?

 

[Peter-]

Solved. Apparently, it's that time of the month when Plesk has a problem with everything.

After abdi's suggestion, I increased my buffers:

#cat /etc/sw-cp-server/config

...

fastcgi_buffers 32 32k; // was 16 16k
fastcgi_buffer_size 64k; // was 32k

...

For a while, plesk admin stopped working. It was because of the timezone. For some dumb reason. It was working before, not working after changing the buffers. I don't see the reason, but ok.

So, next, I had to remove the timezone data.

#yum reinstall tzdata

Next, restart Plesk:
#/etc/init.d/sw-engine restart

Worked.

Waiting for the next server fault.