• If you are still using CentOS 7.9, it's time to convert to Alma 8 with the free centos2alma tool by Plesk or Plesk Migrator. Please let us know your experiences or concerns in this thread:
    CentOS2Alma discussion

Resolved aps-cache task error

Pleskie

Regular Pleskian
Sometimes at night some task that Plesk executes seems to fail.

The error message in panel.log is:

[2016-05-31 04:03:00] ERR [panel] Task failed: id=6, pid=25758, type=aps-cache, error=Connection timed out after 30000 milliseconds

Does anyone know what is causing this error ... and how to solve it? My hosting company doesn't know what it means and what to do about it.

Some days the error message appears, some days it doesn't. Seems quite randomly, but WHEN it appears it is always at night around 04:00.
 
Thanks for your reply UFHH01.

Does this mean that you don't know what the error in my first message means? Is it something you have not seen before?

I rent a VPS and "out of the box" it gives me this error. I haven't modified anything.
 
Hi Pleskie,

Does this mean that you don't know what the error in my first message means? Is it something you have not seen before?
no, this means, that your error message can depends on various reasons, which I don't want to list all, because it might not be YOUR specific cause.
If an error returns mostly each day at the same time, it is most likely, that a daily task causes this error, but as stated above, the error might as well have another cause, depending to the task and the commands for it.

Investigations are normally done based on facts, rather than guessing a root cause, because guessing could push you to the absolute wrong direction and might be a total waste of time. The suggestion to choose another log - level will help you to investigate issues, because the error - log will be more verbose.
 
Thanks ... I'm about to reinstall the server right now, so I won't be able to test your solution (verbose logging) immediately.

I can't find this same error anywhere and that is what bothers me a bit. Last question ... is there a specific Plesk task that runs every night about 4 o'clock? Or is this something the hosting company has scheduled?
 
Hi Pleskie,

Or is this something the hosting company has scheduled?
The hosting company does not configure Plesk - you are! ^^

I wonder why you are re-installing your server, if you only experience issues with a single task, which depends on a connection error. As you already saw in your error message, the task tries to connect to the Application catalog for the daily maintenance ( aps-cache ), but without any verbose output, you can't investigate the connection issues.
 
Last edited by a moderator:
>> The hosting company does not configure Plesk - you are!

Should've been more clear ... I can choose the OS and plesk version based on a preinstalled image. So i thougth somse adjustments had already been made by hosting provider.

>> I wondern why you are re-installing your server, if you only experience issues with a single task

Not only for that one issue. Have been doing some testing the last week to find out how some things work. Now I know most of them so that's why it's time for a clean install :)
 
I have tried the verbose logging ... and I turned on logging queries as well ... wow ... that's a lot of queries :eek:

Anyhow ... on of the messages I found that might be interesting is this one:

[2016-06-06 04:02:27] WARN [panel] Failed to fetch customizations from Partner Central (https://ka.plesk.com:5224): Invalid HTTP response, HTTP error 504

Something seems to go wrong here. Any idea what is causing this?
 
(Thanks, since I'm new here I need to learn to find my way.)

The proposed solution seems to involve a host file with incorrect information or a problem with the firewall. However ... I assume these problems cannot occur with version 12.5 out of the box?
 
Hi Pleskie,

Plesk and people willing to help you don't know, "how" you installed your operating system, depending services and configuration files. If you for example used an image file from your server provider, we can't tell, which sort of settings are done and if you don't provide informations about it ( for example configuration files ), we can't tell if the host file is configured as suggested, or if the firewall blocks a necessary port - the provided article just shows, how you might be able to solve your issue. ;)
 
Thanks ... I do appreciate your help!! The ports that should be openend are opened as we discussed yesterday :)
Strangely enough the error doesn't appear every day. I'll keep an eye on it and maybe will come back later with more information.
 
Well, I've done some testing as described on the link you gave me. Nslookup says:

Non-authoritative answer:
Name: ka.plesk.com
Address: 195.214.233.80

This seems alright.

And also some lines from traceroute:

5 ae1-6.RT.IRX.ZRH.CH.retn.net (87.245.233.230) 16.183 ms 16.173 ms 16.161 ms
6 GW-Parallels.retn.net (87.245.246.185) 19.384 ms 19.408 ms GW-Parallels.retn.net (87.245.246.187) 19.155 ms
7 cx7-ix-dc.parallels.net (195.214.233.34) 19.311 ms 19.324 ms 19.311 ms
8 * * *

Seems there is a connection here as well.

So ... anything else that could be wrong then? More tests I could perform perhaps? Or could it maybe be a coincidental hick-up of the plesk server?
 
YES!!!! I finally have a better log now. I will post the relevant lines. At first the aps-cache task seems ok, but later it seems to fail.

I only post the relevant lines.

This part seems ok:

[2016-06-11 04:02:36] INFO [panel] == Task run APS cache update: /usr/local/psa/bin/sw-engine-pleskrun "/usr/local/psa/admin/plib/DailyMaintainance/task-script.php" UpdateApsCache '--period=daily'
[2016-06-11 04:02:36] INFO [panel] == Begin APS cache update
[2016-06-11 04:02:37] DEBUG [util_exec] [xxxxx] Starting: php /usr/local/psa/admin/bin/php '/usr/local/psa/admin/plib/scripts/task-async-executor.php' '-task-id' '12'
[2016-06-11 04:02:37] DEBUG [util_exec] [xxxxx] Finished in 0.00414s, Result: TRUE
[2016-06-11 04:02:37] DEBUG [panel] New update aps cache task was registered
[2016-06-11 04:02:37] INFO [panel] == End APS cache update
[2016-06-11 04:02:37] INFO [panel] == Task finish APS cache update

At the end it fails:
[2016-06-11 04:02:41] INFO [panel] Finish: Daily Maintainance
[2016-06-11 04:03:16] ERR [panel] Task failed: id=12, pid=29231, type=aps-cache, error=Operation too slow. Less than 500 bytes/sec transferred the last 30 seconds
[2016-06-11 04:03:16] DEBUG [panel] PleskFatalException: Operation too slow. Less than 500 bytes/sec transferred the last 30 seconds
file: /usr/local/psa/admin/plib/Aps/Catalog/Feed/Reader.php
line: 270
code: 0
trace: #0 /usr/local/psa/admin/plib/Aps/Catalog/Feed/Reader.php(50): Aps_Catalog_Feed_Reader->getContent(string 'http://apscatalog.com/all-app/', boolean true)
#1 /usr/local/psa/admin/plib/Smb/Plesk/Bridge/Aps/Catalog.php(116): Aps_Catalog_Feed_Reader->fetchRepositoryIndex()
#2 /usr/local/psa/admin/plib/Aps/Catalog/Cache/Task.php(12): Smb_Plesk_Bridge_Aps_Catalog::updateAvailableApsCatalogCache()
#3 /usr/local/psa/admin/plib/Db/Table/Broker/LongTasks.php(262): Aps_Catalog_Cache_Task->run()
#4 /usr/local/psa/admin/plib/Db/Table/Broker/LongTasks.php(180): Db_Table_Broker_LongTasks->_syncStart(object of type Db_Table_Row_LongTask)
#5 /usr/local/psa/admin/plib/Db/Table/Broker/LongTasks.php(112): Db_Table_Broker_LongTasks->_startQueue(object of type Db_Table_Row_LongTask, boolean true)
#6 /usr/local/psa/admin/plib/Task/Async/Executor.php(52): Db_Table_Broker_LongTasks->start(object of type Db_Table_Row_LongTask, string 'synchronously')
#7 /usr/local/psa/admin/plib/scripts/task-async-executor.php(12): Task_Async_Executor->execute()
As you can see it says "PleskFatalException: Operation too slow. Less than 500 bytes/sec transferred the last 30 seconds".

What is causing this problem? Please someone explain?
 
Hi Pleskie,

your error is really nothing, that you should worry about, because the daily maintenance script can be done in separate tasks as well:

How to rerun daily maintenance tasks separately ( KB - article 114797 )

If one of the tasks doesn't finish with success, there might be several reasons and you already experienced some connection issues AT THE PLESK SERVER SIDE, which you can't control. The root cause in such cases may be heavy loads, daily updates and temporary inaccessibility and in some rare cases, a misconfigured VPS, with small ressources.
If you still desire to finish ALL the daily maintenance tasks without failures, just try to run the tasks separately, as described in the KB - article.

Just to state this clear and without the possibility of interpretations: A newly installed dedicated server ( or VPS ), which has been installed with images is by far not "up-to-date" and needs quite a lot of updates, in your case, the aps-calalog is for example not up-to-date, which is really no big deal. Consider to update your server and it's components first ( with yum or apt-get ), to make sure, that your system components are fully updated, because old components can prevent additional updates/upgrades for third-party components. An old aps - catalog will not result in issues or failures with your Plesk Control Panel - it might just result in installing not the newest version of a desired app from the catalog, which is afterwards "repairable" with a click on the button "update".

Old protocols for example may as well cause your issues and again, pls. update your server components, to resolve this possible root cause.

Last but not least, your network connection might be the root cause, which needs deeper investigations on your server itself and can't be answered with the help of your few provided informations from the Panel log.
 
Thanks for your reply.

Just to be sure ... the updating you are talking about is yum update correct? I already do that regurlarly so I should be good I assume? Can you please confirm?
My Plesk updates are being performed automatically so that should be good as well?

I just ran this command via my terminal:
/usr/local/psa/bin/sw-engine-pleskrun /usr/local/psa/admin/plib/DailyMaintainance/script.php -f UpdateApsCache

I didn't get any error and I don't see any error in panel.log.

I also don't see the error EVERY night, it happens a few times in a week. So does this mean that it is indeed some connection problem that sometimes occurs and I just shouldn't worry about it?
 
Hi Pleskie,

the updating you are talking about is yum update correct? I already do that regurlarly so I should be good I assume? Can you please confirm?

No, sorry.... I can't confirm that, because you didn't provide necessary informations about your operating system.

"yum" is used on RHEL/CentOS - based systems.
"apt-get" ( aptitude ) is used on Debian/Ubuntu - based systems.

My Plesk updates are being performed automatically so that should be good as well?
Not at all. Plesk does not perform "yum" or "apt-get" ( aptitude ) updates. The server administrator itself is responsible for the operating system depending updates/upgrades and it's services.

I also don't see the error EVERY night, it happens a few times in a week. So does this mean that it is indeed some connection problem that sometimes occurs and I just shouldn't worry about it?
Pls. answer this question BY YOURSELF: It is a DAILY MAINTENANCE SCRIPT, so this script is supposed to run EVERY DAY. If you experience errors only on several days AND NOT EACH DAY, then you can be sure, that the script did its job without failures, if you don't see possible errors in the daily error - log.


Nevertheless, IF you experience returning issues, you should still investigate the root cause ( network issues ?!? ), because it is very rare, that the root cause are issues on the Plesk server site.
 
@Pleskie,

I normally agree with what @UFHH01 states and his latest response is not an exception.

However, there is one subtle nuance: it is not rare that issues are present on the Plesk (licensing) server side.

To illustrate, I have 100% identical servers, some of them have license update related issues, others do not.

As another illustration, I have one server that has "APS update" issues (while other parts of license checks and updates are fine).


Nevertheless, the general conclusion (and personal advice to you) is that you should not be worried about each error message.

This particular error message (license related) completely harmless and can be safely ignored. Just as I and @UFHH01 have repeated many times.

Ciao
 
>> No, sorry.... I can't confirm that, because you didn't provide necessary informations about your operating system.

Sorry ... I was talking about CentOS.

>> Not at all. Plesk does not perform "yum" or "apt-get" ( aptitude ) updates.

I understand. I didn't expect it to. I meant the Plesk updates. I have configured Plesk (in the Plesk panel) to automatically update its components. Yum updates I perform myself.

>> If you experience errors ... in the daily error - log.

Yes, o,f course that I understand. I was only asking if my conclusion was right. Since it doesn't happen every night we can isolate the problem to a connection error, be it on my side or on your (Plesk) side ... but how do we find out?

>> Nevertheless, the general conclusion (and personal advice to you) is that you should not be worried about each error message.

These connection errors are actually the only errors I'm experiencing.

>> This particular error message (license related) completely harmless ...

This error message is not license related but has something to do with the apps catalog.
 
@Pleskie

I will react to the your responses to my posts.

>> Nevertheless, the general conclusion (and personal advice to you) is that you should not be worried about each error message.

These connection errors are actually the only errors I'm experiencing.

I know, some or most people do. At this moment, somewhere in the past or the future. And in the past, it always ended well. No worries, mate!

>> This particular error message (license related) completely harmless ...

This error message is not license related but has something to do with the apps catalog.

Actually, you are right, in the sense that I used some short-hand notation that is not entirely correct or wrong: it is about the connection with ka.plesk.com.

Note that despite of the error notification everything should be working on your server, as intended. Even the APS catalog.

Ciao
 
Back
Top