Clusters stuck on pending indefinitely (community edition)

J_
New Contributor II

From yesterday, suddenly clusters do not start and are in the pending state indefinitely (more than 30 minutes). From a previous post, I tried to add 443 port to the firewall but it doesn't help.

In the clusters page, the message says: Finding instances for new nodes, acquiring more instances if necessary

Anonymous
Not applicable

Hello, @Joonkyung Kim​! My name is Piper and I'm a community moderator here. Welcome and thank you for your question. I'm sorry to hear that you're having this difficulty. Let's see if any other members are having this experience and/or if they can offer any solutions. If not, we will get back.

Please let us know if things change.

kburchfiel
New Contributor II

I'm having a similar issue. A cluster that I tried to set up remaining in pending status for about 45 minutes, when I went ahead and deleted it. I'm now trying again with a second cluster.

Hubert-Dudek
Databricks MVP

Hi you need to share your edition and event and driver logs (I had such a problem for example when it can not download library from Microsoft as Microsoft server was down)


My blog: https://databrickster.medium.com/

I think it was an intermittent issue. If you see it happen all the time, then it might be bigger issue. I would recommend to check the status page in case there are any outages during that time https://status.databricks.com/

View solution in original post

Prabakar
Databricks Employee
Databricks Employee

Hi @Joonkyung Kim​ could you please share the event logs of the cluster? It should give you some information on why the cluster did not start.

Reet
New Contributor II

I am also having same issue and there seems to be no outage.....