Unauthorized heartbeat API
Hello,
in our log files I keep finding a recurring log related to a ‘heartbeat’ API.
Has anyone seen this before? I’v e seen the heartbeat implementation in some other apps / sites and I think it’s a useful way to be able to alert for any outages.
However, in our case it seems the place where it’s being called for is not using correct authorization.
These are the logs that keep popping up :
2022-12-21 06:40:31.643 [http-nio-0.0.0.0-4400-exec-8] INFO c.c.i.adapters.web.util.FilterUtil - Returning ‘HTTP 401 Unauthorized’ on request ‘GET /rest/2.0/auth/sessions/heartbeat’ []
If anyone has seen this implemented and is using it , I’d like to hear more about it!
Thijs
thijswaleson
Posted 3 years ago · Edited 1 year ago·Last reply 3 years ago
14 comments
daverichard
·3 years ago · EditedInteresting, the setup and configuration happened a few months before I was brought on, but I do not recall changing that default in any of the meetings I watched. Of course there was a lot of information overload. It is nice to be able to reset the level on that.
Also, that is funny, this morning there were no heartbeat errors (and still are none in our dev logs), but at some point today it started in Prod. I am adding that back to WARN.
Thank you for all of the assistance @arthur.burkhardt !
daverichard
·3 years ago · EditedFortunately, we are not getting these again, I am curious what triggers them to start as we did not modify the log settings, just restarted our instance.
That said I would say a good 80% of our log files are constant
I would love to turn that off as most times I am just simply looking for an error related to a workflow, and before I am even able to get to the log the error is more than 200 lines off screen. It has meant that most times I am looking to review that error I just have the console open on auto-refresh.
daverichard
·3 years ago · EditedFortunately, we are not getting these again, I am curious what triggers them to start as we did not modify the log settings, just restarted our instance.
That said I would say a good 80% of our log files are constant
I would love to turn that off as most times I am just simply looking for an error related to a workflow, and before I am even able to get to the log the error is more than 200 lines off screen. It has meant that most times I am looking to review that error I just have the console open on auto-refresh.
arthurburkhardt
·3 years ago · EditedWell, you just need to add that your configuration

This is how it is defined in our configuration
Edit: BTW, @drichard.beyondtrust.com, this is the OOTB configuration. That means that someone played around with your logback configuration and changed that level
arthurburkhardt
·3 years ago · Edited@tomas.kissik.merck.com @thijs.waleson.heineken.com
We have the same issue, it is VERY annoying.
I opened a support ticket, but they were unable to help me. Here are some extracts of my messages:
You can get rid of the messages by raising those two classes logging level to WARN
tjbrodeur1
·3 years ago · EditedThanks for getting the full class path on these. We had the same issue and decided to do the same silencing, so this helped.
We ran a support ticket also, mostly asking whether there was anything we should be concerned regarding these INFO log messages. The long and short of it is essentially behavior as expected.
Anyway, thanks again for doing the heavy lifting on this.
arthurburkhardt
·3 years ago · Edited@tomas.kissik.merck.com @thijs.waleson.heineken.com
We have the same issue, it is VERY annoying.
I opened a support ticket, but they were unable to help me. Here are some extracts of my messages:
You can get rid of the messages by raising those two classes logging level to WARN
arthurburkhardt
·3 years ago · EditedYou’re welcome! BTW, I juste opened an ideation ticket for this if you want to vote: Ideation Platform – Collibra | Product Resource Center
tjbrodeur1
·3 years ago · Edited@arthur.burkhardt Are you getting these again? We set the Classes you listed to WARN as you described and it seems to have worked as expected, however we appear to be getting INFO level notifications from c.c.i.adapters.web.util.Filters again. Same heartbeat authorization message.
Just curious if you’ve recently seen an uptick in these again.
arthurburkhardt
·3 years ago · Edited@tj.brodeur ahahahah… So glad to hear I’m not the only one being super annoyed with this.
Yes, they renamed the class from
com.collibra.identity.adapters.web.util.FilterUtiltocom.collibra.identity.adapters.web.util.FiltersAnd yes, I’ve been getting tonnes of it
tjbrodeur1
·3 years ago · Edited@arthur.burkhardt Thank you as always. Yes, this is quite frustrating.
Ann Wuyts
·3 years ago · EditedHi Thijs,
The heartbeat references you see in the logs are related to authentication, and not to application moninitoring (confirmed by our SRE team).
GET /rest/2.0/auth/session/heartbeat is used by the UI to poll the backend to check the liveness of the user’s session. It returns 200 if the user session is active. and 401 if the session is expired or doesn’t exist.
https://developer.collibra.com/rest/data-governance#/operations/heartbeat
tomaskissik
·3 years ago · EditedHi Ann,
please do you know if it’s possible to configure logger in console to omit this heartbeats from the logs?
Thank you,
Tomas
Ann Wuyts
·3 years ago · EditedHi Tomas, no sorry. This would be something to check with our Support team.