Hello people. Yesterday I had to choose the vagrant based installation of 11.8.0 version of the K CE edition which went fine to the end. On the front page of the server and under the admin console pointer there is a dead link of it’s docs (http://www.kaltura.org/kaltura-community-edition-kalturace). At the knowledge base the most up to date link points to the gemini version of the pdf (it dates back to 2013). That should not be the current version, is it?. Alternatively is there any specialist up to date reference for the Batch Process Control section?
Nothing much changed in admin console as far as Batch Process Control since Gemini.
If you have specific questions that you could not find answers to in this guide, please ask them and I’ll be happy to answer:)
Alright! Basically I’m on the acquainted phase with the platform. I created a publisher account who successfully submitted the same video subject under two different formats 94MB mp4 and then a 78MB avi. To cut the story short on the first occasion I had several tasks in front of my last submitted task under a different partner id that I suppose is system installed (id: 99 “Template KMC account”. At an earlier stage I forgot to mention that I observed that the timezone of the installed VM was incorrect so I had to sync it to my current local timezone being Europe/Athens and just in case some synchronization or datetime math was the cause of the bottleneck in the BPC queue. So after modifying that and rebooting the VM still there was no change with my media. Aborting the item in front (could not find how to pause and prioritize my submitted media task) the media is still pending on the first step of the first encoding of the plan (i saw this from the publisher’s details at the KMC - view details of the submission and the flavors tab) without any error shown from the entry investigation. And that is the last point where I cannot do any further testing of the platform since the task cannot finalize a complete submission.
P.S.: I can post the mediainfo tool reports of the two video files if required
The upload is still converting, no timeout has interrupted it. I will note down subsystems and logs to check for possible hints on the problem. Also posting a snapshot from the admin console.
Are you certain your batch daemon is running?
What is the output of:
# /etc/init.d/kaltura-batch status
Also, what do you see in /opt/kaltura/log/kaltura_batch.log and /opt/kaltura/log/batch/convert*log?
I am also a bit confused by your description above. Partner 99 is, like you stated, the KMC template account. It is not meant to be used for uploading entries or any other actions, apart from defining the template. When a new partner is created, it inherits partner 99’s set of definitions.
Did you create a new partner from the admin console are using it?
Hello Jess, I mentioned the KMC template user in order to put the initial hello world use case failing into context. Before starting to pick things or examine logs I tried to minimize the chance of extending the complexity of the problem. I deleted the vagrant vm till I have more time to perform a manual installation and review the subsystems involved. If you have any additional hint I will soon retry the vagrant setup and follow your problem solving approach then. The lastlog of the system was full with several repeating blocks of the following log pattern:
Feb 19 13:08:04 host monit[1905]: ‘batch’ process is not running
Feb 19 13:08:04 host monit[1905]: ‘batch’ trying to restart
Feb 19 13:08:04 host monit[1905]: ‘batch’ start: /etc/init.d/kaltura-batch
Feb 19 13:09:04 host monit[1905]: ‘batch’ failed to start (exit status 0) – /etc/init.d/kaltura-batch: Stopping Batch Manager… #012Service Batch Manager not running#011[#033[1;31m FAILED #033[0m]#012Starting Batch Manager… #012Batch Manager started with PID #011[#033[1;32m OK #033[0m]#012
Feb 19 13:09:14 host monit[1905]: ‘host.domain’ cpu system usage of 91.8% matches resource limit [cpu system usage>30.0%]
Feb 19 13:09:14 host monit[1905]: ‘sphinx-populate’ process is not running
Feb 19 13:09:14 host monit[1905]: ‘sphinx-populate’ trying to restart
Feb 19 13:09:14 host monit[1905]: ‘sphinx-populate’ start: /etc/init.d/kaltura-populate
Feb 19 13:10:14 host monit[1905]: ‘sphinx-populate’ failed to start (exit status 0) – /etc/init.d/kaltura-populate: Fri Feb 19 13:09:14 EET 2016#012Starting:#012Service populate isn’t running but stale lock file exists#012Removing stale lock file at /opt/kaltura/log/populate.pid#012/usr/bin/php populateFromLog.php >> /opt/kaltura/log/kaltura_populate.
I browsed for the lock file and at that timepoint it didn’t exist but I didn’t know from where to begin and whether the installation was indeed complete. For instance the monitor tab on the admin console was not showing up at all (despite the fact that all 26 page screens of the installation had only a few warnings).
Your system seems to be in relatively high load but that still does not account to why pretty much none of the daemons seem to be running.
Like I wrote before, start with:
In addition, run:
# kaltlog
from the shell and look at the errors, then, if they’re not clear enough by themselves, look at /opt/kaltura/log/kaltura_api_v3.log and share the relevant lines from there.
Hello Jess, thank you very much for following up the troubleshooting. I will go though a fresh new vagrant installation. The configured RAM for that VM had been 4GB, as for the rest of configuration i modified the vagrant file for a bridged public, dhcp network on a wired connection with a fixed MAC address (in order to get a particular IP assignment). I also removed the domain variable from the admin’s e-mail since it had been different to the specified host (the naming of the variable as kaltura domain was also a bit misleading - maybe kaltura_FQHN) but no other special modification, I think I also changed the root password, had it not been accessed from the same env variable but coupled I can’t be sure. On the upcoming installation I will be very very cautious.