Users suddenly no longer showing in v7.0.2.Ga3 - Showstopper!

[Ashley Pietersen, modifié il y a 7 années.]

Hi, I am evaluating Liferay CE v7.0.2.GA3 for some weeks now and I like it very much, sofar. But now, suddenly, users are not visible anymore under Control_Panel/Users..! Allthough the demo users (3 test users) can still login with their account. This is a showstopper for me unless I can solve this issue. We run Liferay on a VM with Windows2012R2 with 12 mem, 1ooGbyte. Database: MySql (Bitnami Liferay Stack)

I have seen several other posts about this issue, but none gave me a solution other then to rebuild it from scratch. For an evaluation site that is no fun but not directly a showstopper. But for a production site this is a real showstopper. Because I do not know what caused this issue. Makes me scarry....

What I have done sofar:
- issue shows up in all browsers, cleared the caches
-reindex several times - and now reindex seems to keep running for ever
-rebooted several times

Does some one has a solution without starting from scratch?

kind regards
apietersen

[David H Nebinger, modifié il y a 7 années.]

You have to reindex. Go to control panel to Configuration -> Server Administration and choose to reindex all.

[Ashley Pietersen, modifié il y a 7 années.]

Hi David,

Thanks for your response.

I understand and I have done that. But it keeps on running endless - in fact it seems not to be the solution.
Note. This is a almost an empty Liferay v7 portal with 3 users and small amount of documents - it is running for hours to reindex!

kind regards
apietersen

[Ashley Pietersen, modifié il y a 7 années.]

After 24 hours busy with reindexing, users still not visible (accept the admin account) - how long should I wait before its is finished ?!
I think the portal has became useless :-(

[Jorge Díaz, modifié il y a 7 années.]

Did you stop Liferay server during reindex execution?

If you killed server perhaps background tasks become inconsistent,

Execute following queries:

SELECT * FROM BackgroundTask
SELECT * FROM Lock_

And check background tasks status and locks.

[Ashley Pietersen, modifié il y a 7 années.]

Hi Jorge,

Thanks for the repsonse. I ran the query you suggested (in MySQL workbench), this is the result:

'0', 'efe8e5e6-7ff2-8748-d107-b409d37b4c93', '44603', '20116', '0', NULL, '2017-03-21 07:24:01.350000', 'com.liferay.portal.kernel.backgroundtask.BackgroundTaskExecutor', 'com.liferay.portal.search.internal.background.task.ReindexPortalBackgroundTaskExecutor#20116', 'reindex#44602', '0', NULL

Not sure what I can do next...

[Ashley Pietersen, modifié il y a 7 années.]

It looks as if my users are back again, this is what I did:

I ran a query in the workbench of MySql: (bitnami_liferay is the database name that comes with the bitnami Liferay stack)
use bitnami_liferay;
SELECT * FROM BackgroundTask;
I looked in the result list if some tasks where not completed or orphaned, found 2 entries that looked suspisous . I deleted them with the following mysql query:

delete from BackgroundTask where backgroundtaskId=xxxxx;
note: xxxxxx is de ID of that task.
after that I rebooted the server. Still not showing the user after it came up again.

after cleaning the caches etc I reindexed only the index:
Reindex com.liferay.portal.kernel.model.User.

After that my 3 demo users were visible again.
Will test it further and will reindex later on all indexes and see what happens. Keep you posted.

[Rahul Joshi, modifié il y a 6 années.]

I am facing similar issues with Liferay DXP. Some user roles are not able to see documents. Also, if I go to Server Administration->Resources pages when logged in as Admin, it shows issues related to Indexing/Background task:

09:10:48,238 ERROR [http-nio-8080-exec-38][IncludeTag:128] Current URL /group/control_panel/manage?p_p_id=com_liferay_server_admin_web_portlet_ServerAdminPortlet&p_p_lifecycle=0&p_p_state=maximized generates exception: java.lang.NullPointerException
java.lang.NullPointerException
	at com.liferay.portal.kernel.backgroundtask.display.BaseBackgroundTaskDisplay.getBackgroundTaskStatusAttributeLong(BaseBackgroundTaskDisplay.java:132)
	at com.liferay.portal.search.internal.background.task.display.ReindexBackgroundTaskDisplay.getPercentage(ReindexBackgroundTaskDisplay.java:36)
	at com.liferay.portal.kernel.backgroundtask.display.BaseBackgroundTaskDisplay.hasPercentage(BaseBackgroundTaskDisplay.java:77)

[Ashley Pietersen, modifié il y a 6 années.]

Hi Rahul,

I suggest you to look for orphaned background tasks as was mentioned before in this post.
Besides I noticed that when I click to run indexing for all dexes it has a bigger change to hang. If you it one by one, you will proable see that some do run much longer then the other. I suppose that the timing and queing of the jobs might be the problem.

succes!

[Rahul Joshi, modifié il y a 6 années.]

Hi Ashley,

Can you please help me understand how to figure out the orphaned jobs?
So, you suggest that I don't go with the All option and instead run indexes one by one?
Additionally, as mentioned in the previous message, the resources page is not accessible right now and shows a Null Pointer Exception. Is there a way to get past it so that I can run the indexes?

[Ashley Pietersen, modifié il y a 6 années.]

Hi Rahul,

To look for the orphaned background tasks you need access to your database environment. I use MYSQL on my Liferay instance, but the same is true for any other database platform. If you have access and enough rights to your database and run the query something like this:

(Note: you can run these queries from you database manager, normally)

use <Liferay_database_name>;
SELECT * FROM BackgroundTask;

notes: above query this does not change or harm your database, it just read it and show potential BackgroundTask s


and when you are sure you have located the right one:

delete from BackgroundTask where backgroundtaskId=xxxxx;

note: the exact punctuation (here the ";" at the end of every line) of the query depends on the database you use

hope this will help

[Randy Schwartz, modifié il y a 4 années.]

Thanks for sharing! Same issue in 2019 -> DXP 7.2
It makes no difference whether the default search (elastic) or clustered search (solr / elastic) is in use. To many service restarts at the wrong time and indexing never ends.

Solved by manipulating the database (MS SQL):
-- List tables
SELECT * FROM BackgroundTask
SELECT * FROM Lock_
-- Remove all entries in lock table, else the backgroundtask appeared again
delete from lock_ where userid=0
-- Removing all not completed schedules
delete from backgroundtask where status=0
Afterwards I restarted all liferay nodes