User is able to create the report, the issues comes when he is trying to save the report. The error message we are encounter while saving the report is given below.
Could not save the document to the repository for the following reason: repo_proxy 30 DocumentFacade::uploadBlob - Query execute has failed : Error occurred while attempting to reconnect to CMS : Not a valid logon token. (FWB 00003) (hr=#0x80042a70) (WIS 30567)
Configuration is:
BOE XI 3.1 SP2 (Windows 2003 SP2)
Apache Tomcat (Java based)
Please help me to fix this error. Your help is highly appreciated.
We just started having this come up on our servers (Developement, QA and Production) all within the same week.
Furthermore it’s not just WEBI so restarting the web service probably won’t fix the problem. We have the same issue in QAAWS.
I saw that ADAPT notice but SAP said it’s not coming out until SP4 which has no date yet.
I find it odd that we’ve been running with the most recent service/fixpack for the last few months (Upgraded to FP 2.7 the week it was released) and only now am having this issue.
There has to be some sort of external cause… but SAP doesn’t seem able to tell me where to look, only that it’s being worked on. THey couldn’t even tell me if the issue happens over time, memory based or anything that might give me a clue to mitigate the problem.
Having an ad-hoc reporting system where users are unable to save reports pretty much defeats the purpose, so I don’t know why this isn’t a higher priority for SAP.
Getting very frustrated here with SAP’s resolution center.
If anyone else has any idea of what the cause might be (I’m guessing it’s network releated and not just BOBJ’s fault) please post your thoughts… I’ll try anything at this point.
I was having that issue with scheduled reports a couple months ago (we were also on FP 2.7). I restarted all of our servers, and it went away. I can’t say for sure that that is the fix, but it worked for us.
When you say “restarted all of our servers”, are you refering to Business Objects internal CMS servers (WebIntelligenceServer, JobServers, etc…) or the physical hardware/OS itself?
Our hardware is spread across 3 machines, an APP server, the TomCat on it’s on Web Server and our SQL repository on a 3rd box. I can restart them all I suppose… perhaps it’s just a network sync’ing issue of some sort…
Folks, I am pasting the info from SAP NOTE 1432731 - date 07/06/2011
===============
Cause
The CMS prematurely deletes the Web Intelligence Report servers logon token information.
This issue is being tracked as Problem Report ADAPT01374806 (CMS prematurely deletes all Webi server’s server logon token after it crashes)
Resolution
This issue is scheduled to be fixed in Service Pack 5 for Business Objects Enterprise XI 3.1
Workaround for refreshing on demand:
Restart the Web Intelligence Report servers so they generate a new server logon token
Workaround for Scheduling:
After restarting the CMS/SIA, restart the Adaptive Job Server(s) and WebI Processing Server(s).
Thanks Saddat.
Glad to hear it’s being corrected(and not just left for the XIr4 folks).
Our solution ended up being setting up a windows service restart every night for the SIA services (BO’s server intelligence agent) . That seemed to drastically reduce the number of occurances in the short term.
We’re playing with this now as well. We were given a couple things to try that I’m still messing around with, but an interesting piece of news from my support person was that he thought SP5 would probably be arriving sometime around October 2011.
This bug missed the boat when SP4 shipped, and it’s too complicated and touches too many files to go out as a fixpack (or hotfix or limited availablility fix or whatever the cool kids are calling it these days) so we have to wait for the next full Service Pack.
Oh, the two things I have to fiddle with were shortening a timeout and adding a second CMS instance. I wanted to see if the second CMS would preserve the tokens even if the first wiped out, and he wanted me to change one parameter for WebI Processing Server from the CMC. “Please change the #Timeout Before Recycling# value from the default 1200 seconds to 120 seconds. After making the change restart the WebI Processing Servers.” I’m guessing by expiring the tokens quickly they get written back to the repository instead of hanging around in the CMS memory where they get lost in a crash (?). Worth a shot I suppose.
(BOB member since 2005-07-18)
(BOB member since 2009-08-19)
(BOB member since 2005-03-18)
(BOB member since 2008-12-18)