Camunda 8 Self-Managed Elasticsearch not starting

Camunda 8 Topics
1

Hello,
i am currently trying to run a test installation of camunda 8 as described here: Camunda 8 Helm installation | Camunda 8 Docs
I haven’t changed anything in the default config.

After successfully running the helm install, my k3s cluster looks like this:

image001
image0011847×541 53.7 KB

After checking all the logs it seems to be waiting for Elasticsearch to start, so the other components can start as well.

Elasticsearch does not throw any error in its logs just a warning as follows that keeps repeating:

[2024-10-29T13:42:53,616][WARN ][o.e.c.c.ClusterFormationFailureHelper] [camunda-elasticsearch-master-0] master not discovered yet, this node has not previously joined a bootstrapped cluster, and this node must discover master-eligible nodes [camunda-elasticsearch-master-0, camunda-elasticsearch-master-1, camunda-elasticsearch-master-2] to bootstrap a cluster: have discovered [{camunda-elasticsearch-master-0}{ljA1WcI5SV6tR_o2VdYDVw}{vVU6Fk_yQvSOCWWy1tFYHA}{camunda-elasticsearch-master-0}{camunda-elasticsearch-master-0.camunda-elasticsearch-master-hl.default.svc.cluster.local}{10.42.0.41:9300}{cdfhilmrstw}{8.15.2}{7000099-8512000}]; discovery will continue using from hosts providers and [{camunda-elasticsearch-master-0}{ljA1WcI5SV6tR_o2VdYDVw}{vVU6Fk_yQvSOCWWy1tFYHA}{camunda-elasticsearch-master-0}{camunda-elasticsearch-master-0.camunda-elasticsearch-master-hl.default.svc.cluster.local}{10.42.0.41:9300}{cdfhilmrstw}{8.15.2}{7000099-8512000}] from last-known cluster state; node term 0, last-accepted version 0 in term 0; for troubleshooting guidance, see Troubleshooting discovery | Elasticsearch Guide [8.15] | Elastic

I was looking all day to find a solution but havent found anything usefull so far, maybe one of you guys has an idea?

Greetings

2

Make sure you have sufficient CPU/RAM allocated. Are you running on VM?

3

Yes i am, will try to allocate more CPU/RAM tomorow but haven’t seen any log or event message indicating this to be the problem. CPU and RAM chill at like 5%

4

What’s your K3s cluster configuration? Are you running on single node or multiple node?

5

Same as camunda default install single node like here: Quick-Start Guide | K3s

6

The link you are following is for helm chart on multiple node. In your scenario you are using k3s cluster, single node, please refer kind chart, which is single node Elasticsearch.

7

If you would like to tryout Camunda 8, you can try with C8Run available on windows and unix. It’s simplify the process to get started.

8

Hi,
i knew about the C8Run but this was a test to see how we would run camunda in our cluster. The kind values worked and are better suited anyways, because of the smaller environment. Thanks for that.

If somebody can explain to me why it wouldn’t work with 3 elasticsearch pods i would appreciate it so. I generally like to understand why something fails.

9

Keep in mind that the Kind values are not intended for use in production.

Your elasticsearch pod logs should tell you why it didn’t start up. Pod 2 appears to have been crashing looking for the leader, so something didn’t seem to be happy about EleasticSearch pod 1.

10

Yea i know it’s just for internal dev work for now so they are fine.
As i said the logs didn’t show anything of use except the message above and this message showed up on all of them. It starts one of them with this message leaving the other 2 pending and never reaches ready state. When i kill the one running the next one spins up with the same problem and it creates a new one thats pending.

11

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.