Web Content is temporarily unavailable

[Edwin Arévalo, modificado hace 12 años.]

Hi Experts ...

Before you ask, a brief introduction of the environment.

Description:
* Liferay 6.5 community edition
* Server glassfish 1.3 using cluster configuration.
* Liferay configured with 5 instances and each instance has a portal with lots of content.
* The weight of the application server and the 5 sites is about 3 GB.
question:
* When trying to create new content using the "web content portlet" is present an error with the following message: "Web Content is temporarily unavailable". We executed the following activities to try to solve the problem:
Garbage collection
Clearing VM caches
Clearing caches across the cluster
Clearing database caches
Reindex all search indexes
Finally you restart the domain but does not fix the problem, anyone know how to fix it?

In advance thank you very much for responding.

[David H Nebinger, modificado hace 12 años.]

There should be a corresponding error in the console indicating what the failure was. There's no way to diagnose from what you've provided.

[Hitoshi Ozawa, modificado hace 12 años.]

* When trying to create new content using the "web content portlet" is present an error with the following message: "Web Content is temporarily unavailable".

Are you getting this every time or just sometime?

Liferay 6.5 community edition

You mean Liferay 6.0.5. That release had some GUI bugs. It may help if you upgrade to Liferay 6.0.6.

[Alejandro López, modificado hace 12 años.]

Hi Edwin!!

Finally did you get any solution? I'm suffering similar issue.

Greetings

[Hitoshi Ozawa, modificado hace 12 años.]

There should be an error message in the console explaining the reason. Please attach the console log.

[Alejandro López, modificado hace 11 años.]

Hi Hitoshi.

It's strange, but I don't see any java exception at console explaining the reason...

Greetings.

[Hitoshi Ozawa, modificado hace 11 años.]

How about in the logs (tomcat-6.0.26\logs)?

[Alejandro López, modificado hace 11 años.]

Hi.

I'm using Apache and Weblogic application server. I don't see any exception, nothing strange in apache or weblogic log.

Sometimes publishing this content, cause that error, other times another error ("Another user has edited content while you are editing... "). Other times, edit of this content is OK...

The editor of this content has dropped some images into WYSEWIG causing long html for the image (<img src="data:image..[a long code for the image] ..." /> ")

At the end of the html of this content theres a strange javascript:

<script id="dstb-id" language="javascript">
if(typeof(dstb)!= "undefined"){ dstb();}</script>

But this script is on another contents that are ok. I think it's some index problem but I don't know...
It's a mistery.

Greetings.

[Hitoshi Ozawa, modificado hace 11 años.]

Web Content has some character restriction. Best to stick only with alphanumeric unless you customize.

You're also entering some scripts which may be the source of the error. In most of the cases I've seen, the problem is (beside the limitations) with the content and scripts rather then liferay.

[Edwin Arévalo, modificado hace 11 años.]

Thanks to each of you to answer this question, the problem appears to reindex portlet web-content, the solution may be to restore a database backup or delete the index that ocaciona the problem. Greetings.

[Hitoshi Ozawa, modificado hace 11 años.]

Well, if you tried using liferay's default search, you'll soon find that it isn't too good. It's better to replace it with something different.

[Alejandro López, modificado hace 11 años.]

Hitoshi Ozawa:

Well, if you tried using liferay's default search, you'll soon find that it isn't too good. It's better to replace it with something different.

Finally the problem was that my tablespace were at 94% of his total capacity. We increase the tablespace and the problem dissapears.
Thanks to all!

[Hitoshi Ozawa, modificado hace 11 años.]

Finally the problem was that my tablespace were at 94% of his total capacity. We increase the tablespace and the problem dissapears.

Does this mean you've changed lucene to use database instead of the default file system?

[Alejandro López, modificado hace 11 años.]

Hitoshi Ozawa:

Finally the problem was that my tablespace were at 94% of his total capacity. We increase the tablespace and the problem dissapears.

Does this mean you've changed lucene to use database instead of the default file system?

Hi.

When this error appears, I have indexes on Database, and a tablespace capacity 100MB. Trying to solve the problem, I put the indexes on file but this error persists.
Now I have indexes on file and increase tablespace capacity. Now the problem has disappear.

Greetings.

[Hitoshi Ozawa, modificado hace 11 años.]

It's nice to hear that you've got it working. Just for my record in case similar question appears again, did you restart your liferay after changing portal-ext.properties? Lucene shouldn't be using the database set up lucene if file system is being used.

[Alejandro López, modificado hace 11 años.]

Hitoshi Ozawa:

It's nice to hear that you've got it working. Just for my record in case similar question appears again, did you restart your liferay after changing portal-ext.properties? Lucene shouldn't be using the database set up lucene if file system is being used.

Yes, I restart Liferay putting indexes on file. I think the 2 liferay tables "Lucene_0" and "Lucene_10132" have disappeared making this, but I don't know because I don't have access to the database, and I have to do another things.

But, after restart liferay (with indexes on file), the problem wasn't solved. Maybe we should delete some temporal tables... don't know.

Only when increase tablespace of database the problem was solved.

pd: sorry about my english

[Hitoshi Ozawa, modificado hace 11 años.]

Thanks Alejandro, I'll keep increase tablespace size as one of the possible solutions when portlet is not working correctly.

[Alejandro López, modificado hace 11 años.]

You are welcome!

Take a look on why Oracle error doesn't show at application server log.

Greetings

pd: I've checked "LUCENE_0" and "LUCENE_10132" have disappeared after putting indexes on file.

[Alejandro López, modificado hace 11 años.]

Edwin Arévalo:

Thanks to each of you to answer this question, the problem appears to reindex portlet web-content, the solution may be to restore a database backup or delete the index that ocaciona the problem. Greetings.

Edwin, ¿tenias los indices de Lucene en BD cuando te sucedió esto?
----------------------------------------------------------------------------------------------------------------
Edwin, ¿did You have Lucene indexes on BBDD when you found this issue?