Post Fixpack 10 - Error 1113 (No mapping for the Unicode character exists in the target multi-byte code page.) in TranscodeU8ToU16

We recently deployed fixpack 10 and now see this message logged for most of our top level relays. Before this we only had seen it for one top level relay.

Here is what is on IBM site:
https://www-01.ibm.com/support/docview.wss?uid=swg1IJ05780

And these are messages I am seeing in besrelay.log on root
Uncaught exception in plugin ClientRegister with client xx.xx.xx.xx: WindowsCharacterSetResult: Error 1113 (No mapping for the Unicode character exists in the target multi-byte code page.) in TranscodeU8ToU16

Any suggestions ?

Thx

I had a few cases of that as well, and found that some of my clients had corrupt client settings in the Registry. Take a look at these threads and see if they help.

In my case, these errors started on a few endpoints after upgrading them from 9.5.5 to 9.5.8. I opened up a case with IBM (TS000130487) which resulted in APAR IJ05780 (identified a failure in transcoding the additional data that 9.5.8 collects to prevent computer duplication). I think that fix has made it into 9.5.10 (but not sure).

The troubleshooting process is to “walk” the IP address to the client. In other words, the IP shown on your top level relay (TLR) will be one of your child relays, the child relay will have the same error with another IP. That IP will either be another child relay or the client itself. What is worse, is that the client will jump from relay to relay when it has this symptom, so even a single problem client can cause many log entries on the TLR.

In my case at least, once finding the problem client, I downgraded them to 9.5.5 and the issue was gone.

Thanks AlexaVontess, doesn’t that mean that there is a bug in the endpoint after 9.5.5 or endpoints became sensitive to codepages - would be nice to have an ignore option Downgrading clients is not an option for us unfortunately.

Thanks also JasonWalker, I did see mention of those threads and started looking for corrupted registry settings but didn’t find any. I didn’t think that results from a property could also generate this message and will relook our properties but that could take a long time.

Thx again for your input.

It is a bug, hense the APAR. I don’t see why downgrading the few clients that may have this issue is not an option, but that’s your business. :yum:

Is IBM aware that this was not fixed in FP10?

Thx

Bump…

We are getting this now too. Any thoughts HCL?

We identified another code issue other than APAR IJ05780, always in the computer registration process and always related to clients reporting characters the BigFix Server is not able to handle: it is tracked with APAR IJ09791 (https://www-01.ibm.com/support/docview.wss?uid=swg1IJ09791) and it has been released with 9.5.11.

I can say that 9.5.10 resolved it on one of our system, but another one is still having the issue, although it is sporadic (it connects most of the time where as before, it would simply try to connect to each relay and fail). In short, it isn’t fully fixed.

I am not seeing these messages any longer in 9.5.11. I will continue to monitor our relays for it though.