SAP Data Hub Trial @GCP no more reachable (host site https://sapdatahubtrial/)

[yu_chen_10]

Hi,

the host site of the Data Hub (https://sapdatahubtrial/) is now unreachable, while yesterday it worked. Has anybody an idea for the reason or how to fix it?

I'm wondering why the trial version is so unstable.

Thanks a lot, Yu

[former_member601462]

Hello Thosten,

I am having a similar problem. The Error message is the same as in the first screenshot "Error: Server Error".

It was working a couple days and now https://sapdatahubtrial/ is not reachable. The services shown in SAP CAL seem to work all fine. Only one warning: "There are no available nodes for LoadBalancer service datahub/vora-tx-coordinator-ext-lb-internal"

Is this the reason why the web access https://sapdatahubtrial/ is not reachable?

How to fix that?

Thanks a lot.

Enrico

PS: The GCP console is showing the following:

[schneidertho · Accepted solution]

Good to hear that it works again, Yu.

If the issue comes back, let me know. It's expected that after a restart it takes a moment for all services to be up and running again. Afterwards all "should" work. (Of course) if a VM goes down, K8S will try to reschedule affected pods. Since the trial environment is quite small, you might run into a situation that K8S does not find enough resources on the remaining VMs / nodes. And yes, in that case it might take a while for the system to heal (until K8S has replaced the VM / node)

Cheers
Thorsten

[yu_chen_10]

Hi Thorsten,

we have the problem again, but this time very similiar to the problem of Enrico.

The trial edition instance was built up about two weeks ago. Last week the instance worked well, we have even connected it to our on premise HANA DB. But this week it works not any more.

We could see the same warning like the one of Enrico at the begining. And after a while there was no error or warning more in Data Hub. But the https://sapdatahubtrial/ was not reachable.

We could observe there were some GKE errors in GCP:

Any idea?

Thanks a lot and kind regards

Yu

[schneidertho]

Hi,

hard to say from the screenshots. Does this happen after you have suspended and resumed the solution instance?
And do the "traffic lights" turn green when you wait a while after the resume to allow all services to boot?

If you did NOT suspend/resume, please try once to see if that helps.

If it does not help, I cannot do much without being able to look at the cluster (ideally with kubectl). I don't think it is a general issue, since so far I have not received other complaints.

Are you able to give me access on Friday for 1-2 hours? If you prefer to share logon details via a customer ticket instead of mail, then please create one on component EIM-DH and let me know when you have done. I will pick it up from there.

Cheers
Thorsten

[yu_chen_10]

Hi Thorsten,

thanks a lot for the answer and suggestion.

Yes, the error happened after I have suspended and resumed the instance 3 or 4 days ago. The traffic light didn't get green even after I have waited a long while. Maybe that was some problem with the GCP site?

But as yesterday and today I tried/resumed, the instance is working again. It's interesting, maybe the selfhealing feature of the Ku8 has worked. 🙂

Anyway, I can continue with my work. Thanks.

Yu