ClusteringBootstrap - Cluster started, name: null

[rids01]

We copied one test enterprise environment (ACS-1) (ACS file system + DB) to another test environment (ACS-2). When we attempt to start up ACS-2, it starts up in community mode and we have the following error in our log before we get the SEVERE error:

2021-10-01 08:14:37,477 INFO [cluster.core.ClusteringBootstrap] [localhost-startStop-1] Cluster started, name: null
2021-10-01 08:14:37,480 ERROR [web.context.ContextLoader] [localhost-startStop-1] Context initialization failed
java.lang.NullPointerException
at org.alfresco.enterprise.repo.cluster.core.ClusterServiceImpl.getActiveMembers(ClusterServiceImpl.java:174)

Name:null seems very concerning as it has an assigned value in our other test environments I just have no idea where to start or what to look at.

Any suggestions?

[cristinamr]

Hi @rids01 !

You are asking for an enterprise version: you should contact Alfresco support directly.

But, as far as I remember when I've configured a cluster in the past, I didn't have that problem. Did you follow this steps? Which steps did you perform?

Cheers,

Cristina.

[rids01]

Thanks Christina.

Yes I did contact Alfresco support but it's been a few days and have been getting impatient so I thought I would try the forum. I will escalate at support.

And yes I had gone thru the clustering steps on that link but thank you for sending.

I will contact Alfresco support again.

Thanks,
Aaron

[cristinamr]

I understand your issue... Maybe @amanda_roberts can help you to reach the support faster.

[angelborroy]

According to documentation, the cluster is started creating a cluster name from database

https://docs.alfresco.com/content-services/6.1/admin/cluster/#starting-the-repository-server-cluster

Are you starting only one member? Are you trying to start all the nodes of the cluster at the same time?

[rids01]

Hi,

We are only starting one member but maybe because this started with us copying one test DB to another we left bad names in the database and didn't clean up properly prior to startup. 

Is there a table I could look at to check? Or somewhere in the admin console?

[angelborroy]

There should be some row in ALF_NODE of type descriptor that includes the uuid field as Cluster identifier.

In my local deployment:

select * from alf_node where id=7;
 id | version | store_id |                 uuid                 | transaction_id | type_qname_id | locale_id | acl_id |
----+---------+----------+--------------------------------------+----------------+---------------+-----------+--------+
  7 |       4 |        2 | 2c062b6f-889d-40ee-ad5a-d681d60f83b6 |             14 |            12 |         2 |      4 | 

And 12 means descriptor:

select * from alf_qname where id=12;
 id | version | ns_id | local_name
----+---------+-------+------------
 12 |       0 |     1 | descriptor

You may check if that row exists in your database.

[rids01 · Accepted solution]

Just to close the loop on this one, the cause of our issue is when we copied one test environment to another. We believe we started up the new environment and it started acting like a member of the cloned environment's cluster. 

As per the caution box in this section (https://docs.alfresco.com/content-services/6.0/admin/backup-restore/#back-up-and-restore-repository), this caused corruption. 

To resolve, we restored from back-ups and will be looking into either turning off clustering or doing the network isolating mentioned in the caution box. 

Thanks to all that provided some suggestions.