Sync LDAP Utility Error - Trouble Creating Report
When using the Sync LDAP Utility in Cascade 7.0, we get the following error message
"An error occurred: null"
From what we can tell the actual sync occurs, but looking at the logs it seems that Cascade it is having trouble creating and emailing the report.
"2012-05-29 11:24:47,581 ERROR [LdapServiceImpl] {User: system, id: not specified, type: not specified} Error occurred during Velocity generation of LDAP import report: java.lang.NullPointerException"
and
"2012-05-29 11:24:48,489 ERROR [ThreadedEmailQueue] Could not send email due to error: java.lang.IllegalArgumentException: Text must not be null"
This is all new since our upgrade to Cascade 7.0 late last week. I'm including today's full log file. I'm seeing errors regarding PublishSets as well, though those seem intermittent and unrelated to this issue.
- cascade.log.zip 6.5 KB
Showing the most recent page of comments. View the first page
Comments are currently closed for this discussion. You can start a new one.
32 Posted by erikg on 31 May, 2012 04:11 PM
This may all be a moot point. After the restart last night manual syncing of LDAP was successful without errors and I was emailed a proper report. Reports for scheduled publishes are now appearing correctly too.
I'm still seeing errors in todays log file, but they're different than before. I'm not sure what to make of them or if they're serious.
Was this a one time glitch or something I should keep my eye out for? When I installed I had restarted a couple times to install the Docx Publish Trigger and such. Not sure why this one made a difference. I don't restart Cascade all too often, though.
Support Staff 33 Posted by Bradley Wagner on 31 May, 2012 04:47 PM
Erik, it doesn't surprise me that restarting fixed the problem but I'm still not exactly sure what the issue was. From what we could tell it looked like some kind of class loading problem, library incompatibility, or Java issue. Basically, an error was happening that should have been impossible given the version of Java you were using and the libraries that we ship with Cascade. That's why we were having you check to see if there were some other libraries getting loaded from one of the lib directories or if you were using a non-standard version of Java.
Let's definitely keep an eye out for it going forward and re-open this ticket should you run into it again.
The logs do contain some unrelated errors with database publishing. Is this something you're actively using? If not, it could be some stale data in the database?
34 Posted by erikg on 31 May, 2012 05:03 PM
This comment was split into a new discussion: Errors in log file from database publishing
Sounds good. I'll let you know if it happens again.
We do actively use database publishing for parts of our site. Perhaps I need to flush the external database and republish. I just can't tell if the problem is on the Cascade side or the receiving database.
Thanks for your help!
Support Staff 35 Posted by Bradley Wagner on 31 May, 2012 05:07 PM
Thanks, I just opened a new ticket for that issue and am going to close this one out.
Bradley Wagner closed this discussion on 31 May, 2012 05:07 PM.
erikg re-opened this discussion on 05 Jun, 2012 05:07 PM
36 Posted by erikg on 05 Jun, 2012 05:07 PM
I'm sorry to say that this problem has resurfaced. Just noticed today that LDAP is giving errors once more and publish reports are no longer showing correctly. At no time has Cascade been restarted since you had me do this last time. This is a spontaneous recurrence. The discussion was closed so I cannot attach a new log file, but if you re-open it I will send along a new one. I've already re-enabled LDAP debugging on our server just in case.
Erik
Support Staff 37 Posted by Bradley Wagner on 05 Jun, 2012 05:11 PM
Thanks Erik. Your reply automatically re-opened the discussion so you should be able to attach the files now. Sorry to hear this re-surfaced.
find .from the main Cascade folder and include the output as an attachment?We're looking for any possible differences in libraries, files, plugins -- anything that could help us explain why you're running into this.
Thanks!
38 Posted by erikg on 05 Jun, 2012 05:26 PM
Here's today's log file. LDAP Debugging was only turned on a bit ago, but I had to do a couple syncs since then so hopefully it'll have the info you need.
The only addition we make, as I mentioned before, is the Docx Trigger. That involves the inclusion of a docx directory containing associated material at
tomcat/webapps/ROOT/WEB-INF/classes/com/cms/publish/client. Beyond that it is a standard install. I can send you that material if you'd like. I got it from Rahul Bhagnari who created the trigger for us.The only other thing I can think to mention is that we use an Apache worker agent configuration to present Cascade via SSL. We do this to avoid altering Cascade's tomcat configuration and to present a clean url to end users (by not having to include the port information). I can't imagine if that would make a difference, though. Thought I'd mention it just in case.
Support Staff 39 Posted by Bradley Wagner on 06 Jun, 2012 05:22 PM
Very odd. We're not seeing the same caused by exceptions this time as we did last time that let us track the problem down to a specific line number.
I'm going to do some more research into this issue and get back to you.
40 Posted by erikg on 06 Jun, 2012 05:51 PM
No problem. Last night I bounced the server again to get reporting working for us. We're fine for the time being.
On an additional note, late yesterday, before I bounced the server, I noticed that our weekly LDAP sync happened and I received a properly formatted email as a result. The trouble seemed to be constrained to the manual syncs and publish reports. Hope this helps.
Support Staff 41 Posted by Bradley Wagner on 06 Jun, 2012 09:46 PM
Thank for that information Erik. It's beginning to sound like a race condition and possibly the result of a publish report and an LDAP report being generated at around the same time. Definitely a bug in our app, but we're just not sure where yet. We'll keep you posted.
Support Staff 42 Posted by Bradley Wagner on 07 Jun, 2012 03:14 PM
Hi Erik, I've filed this issue here: https://hannonhill.jira.com/browse/CSI-369.
While we don't have a firm grasp on exactly when this issue surfaces, we have what we believe to be a fix for the problem that we will release in Cascade 7.0.1 coming out very soon.
We'd really appreciate if you can download and install this patch and verify that the issue is resolved. There are a lot of other good fixes in 7.0.1 for users running 7.0 so you'd want to upgrade to this version when you can either way.
I still may have you try a very isolated fix for this issue in your environment before upgrading, but I will get back to you on that shortly.
Thanks for your continued patience and we're sorry you're experiencing these problems.
43 Posted by erikg on 07 Jun, 2012 04:03 PM
Sounds good. We can be pretty nimble when it comes to Cascade upgrades, so I'll be able to install quickly.
No problem if you have something for me to do beforehand. Just let me know.
Many thanks!
Support Staff 44 Posted by Bradley Wagner on 15 Jun, 2012 06:39 PM
Hey Erik,
Cascade 7.0.1 was released yesterday with a fix for CSI-369. We'd really appreciate if you can get that installed at some point and verify that it addresses the issues you were running into with LDAP and Publish reports.
Are you still running into that issue in your environment?
Here are the complete release notes for 7.0.1.
Thanks!
45 Posted by erikg on 15 Jun, 2012 11:08 PM
Thanks for letting me know. I'll install the update this weekend and will let you know how things go.
Luckily we have not run into the issue again since the last time I mentioned it. This is our lowest use time for the CMS, however, so I'm not sure what to read into that. Hopefully with the update we'll simply never see the issue again. Cross your fingers.
46 Posted by erikg on 20 Jun, 2012 08:30 PM
I wanted to let you know that I performed the 7.0.1 upgrade last Friday. I have not seen this problem recur so far. So far so good. If it comes back I'll certainly let you know. Thanks so much for your help.
Support Staff 47 Posted by Bradley Wagner on 20 Jun, 2012 09:06 PM
Thanks for the update, Erick. I'm going to close this issue for now. Feel free to reply or comment to re-open it should you run into this again and thanks for your help confirming the fix.
Bradley Wagner closed this discussion on 20 Jun, 2012 09:06 PM.
Bradley Wagner re-opened this discussion on 03 Aug, 2012 07:00 PM
Support Staff 48 Posted by Bradley Wagner on 03 Aug, 2012 07:08 PM
Hi Erick, I noticed that you downloaded Cascade 7.0.3. Sorry if that's creepy :-), we just like to keep track of clients as they access more recent releases. I just wanted to give you a heads up that we actually reverted our fix for CSI-369 (the one for this discussion) in 7.0.3 as it was contributing to a more serious we were seeing with Velocity leaking memory (CSI-413). We will look to address that in a different way in a future patch release.
49 Posted by erikg on 03 Aug, 2012 11:00 PM
Yes, just a little creepy. :) I do appreciate the warning, though. We also reported the the twitter module bug, so it's a trade-off, I suppose. We haven't applied the 7.0.3 patch yet, so I'll consider whether to hold off or live with the consequences. Either way I look forward to the next patch. Thanks again!
Support Staff 50 Posted by Bradley Wagner on 03 Aug, 2012 11:26 PM
I would definitely recommend 7.0.3 if you're using Velocity macros anywhere in your system. If you're using mostly XSLT it will not be as important.
51 Posted by erikg on 04 Aug, 2012 04:05 AM
I went ahead and installed 7.0.3 this evening. I think you're right, the fixes are worth it. I'll let you know if we see a reoccurrence of this particular issue again. Many thanks.
Support Staff 52 Posted by Ryan Griffith on 05 Oct, 2012 06:41 PM
Hi Erik,
I was going over some older discussions and noticed this one is still open. Have you encountered the LDAP sync issue since upgrading?
Please feel free to let us know if you have any other questions.
Thanks.
Ryan Griffith closed this discussion on 18 Oct, 2012 12:21 PM.