After a create cluster with --restore-from pgo df for that cluster gives an error..

[IJOL]

Describe the bug
After a create cluster with --restore-from pgo df for that cluster gives an error..

nacho@hippo14:~/svn/k8s/production$ pgo show cluster genrios
Handling connection for 8444

cluster : genrios (crunchy-postgres-ha:centos7-12.4-4.5.0)
	pod : genrios-6bb4bb8c7b-gcgr7 (Running) on pan.bitvax.com (1/1) (primary)
		pvc: genrios (8Gi)
	resources : Memory: 128Mi
	deployment : genrios
	deployment : genrios-backrest-shared-repo
	service : genrios - ClusterIP (10.43.255.12)
	labels : autofail=true name=genrios pg-cluster-id= pgo-version=4.5.0 workflowid=c6523a79-dd68-4aaf-a3da-eec59bcf84ba crunchy-pgbadger=false deployment-name=genrios pg-cluster=genrios pgouser=admin backrest-storage-type=local crunchy-pgha-scope=genrios crunchy-postgres-exporter=false custom-config=pgo-custom-pg-config pg-pod-anti-affinity= pgo-backrest=true 
nacho@hippo14:~/svn/k8s/production$ pgo df genrios
Handling connection for 8444
Error: 

To Reproduce
Create a cluster with --restore-from

Additional context

It's a completelñy new uncustomized at all operator 4.5.0 install with storage provided by longhorn (BTW check longhorn it is awesome)

[jkatz]

• Are all of the Pods in the cluster up and available?
• Are you using an external WAL PVC or tablespaces?

[IJOL]

Yes and NO for the question, I suspect that the bootstrap pod is being used, and hence de error, nothing apart from normal is a barebones 4.5.0 install, the only thing that we are doing what we weren't doing already is the restore-from thing.. Saludos, -------------------------------------------- Ignacio J.Ortega skype:ignacioj.ortega El dom., 8 nov. 2020 a las 23:05, Jonathan S. Katz (< notifications@github.com>) escribió:

…

- Are all of the Pods in the cluster up and available? - Are you using an external WAL PVC or tablespaces? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#2029 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAXHYV5XYAGV4L4OWU2CZVLSO4IXRANCNFSM4TOLROBA> .

[IJOL]

My suspicion was correct deleting the bootstrap pod did the trick, I now get pgo df from restore-from clusters.. it's fine as a workaround for me I was a little uneasy about that problem, but now it's pretty simple under the morning lights.. ;) something to look at in the upcoming release..

[jfmatth]

I had the exact issue happen to me on MicroK8s. Deleting the bootstrap POD did the trick!

[jkatz]

Thanks for the report. I agree, this sounds like a bug. Treating it as so.