Question

New FME Azure installation has Postgresql authentication error


I have created a new azure installation following this guidance:

https://docs.safe.com/fme/html/FME_Server_Documentation/AdminGuide/Kubernetes/Kubernetes-Deploying-to-AKS.htm

 

When I look at the logs to my fmeServer Postsql service from powershell

kubectl logs fmeserver-2019-1590072608-postgresql-0

I get the following with authentication errors:

INFO  Initializing postgresql
postgre INFO  ==> No injected postgresql.conf file found. Creating default postgresql.conf file...
postgre INFO  ==> No injected pg_hba.conf file found. Creating default pg_hba.conf file...
postgre INFO  ==> Deploying PostgreSQL from scratch...
postgre INFO  ==> Creating postgres user with unrestricted access...
postgre INFO  ==> Configuring PostgreSQL...
postgre INFO  ==> Configuring replication parameters...
postgre INFO  ==> Configuring permissions for config files...
postgre INFO 
postgre INFO  ########################################################################
postgre INFO   Installation parameters for postgresql:
postgre INFO     Root User: postgres
postgre INFO     Password: **********
postgre INFO   (Passwords are not shown for security reasons)
postgre INFO  ########################################################################
postgre INFO 
nami    INFO  postgresql successfully initialized
[0m[38;5;2mINFO [0m ==> Starting postgresql... 
    Connection matched pg_hba.conf line 95: "host     all             all             0.0.0.0/0               md5"
2020-05-21 14:51:59.420 GMT [157] FATAL:  password authentication failed for user "fmeserver"
2020-05-21 14:51:59.420 GMT [157] DETAIL:  Role "fmeserver" does not exist.
    Connection matched pg_hba.conf line 95: "host     all             all             0.0.0.0/0               md5"
2020-05-21 14:51:59.438 GMT [158] FATAL:  password authentication failed for user "fmeserver"
2020-05-21 14:51:59.438 GMT [158] DETAIL:  Role "fmeserver" does not exist.
    Connection matched pg_hba.conf line 95: "host     all             all             0.0.0.0/0               md5"
2020-05-21 14:52:00.221 GMT [161] FATAL:  database "fmeserver" does not exist
2020-05-21 14:52:00.491 GMT [163] ERROR:  relation "fme_node" does not exist at character 15
2020-05-21 14:52:00.491 GMT [163] STATEMENT:  SELECT * from fme_node
2020-05-21 14:52:01.275 GMT [176] ERROR:  relation "fme_node" does not exist at character 15
2020-05-21 14:52:01.275 GMT [176] STATEMENT:  SELECT * from fme_node
2020-05-21 14:52:01.540 GMT [181] ERROR:  relation "fme_node" does not exist at character 15
2020-05-21 14:52:01.540 GMT [181] STATEMENT:  SELECT * from fme_node

 

I cannot connect to the web portal to the Fme server it appears to time out.

Any ideas why this error occures?


2 replies

Badge

Hi @bknightos,

Sorry to hear you run into this issue!

Can you confirm that you are using the default example in step 4 of the doc you linked?

It looks like the password authentication is failing, but if you didn't change the default pw for the DB this is unexpected and not what we've seen in our tests.

 

 

If you didn't change any of the default settings please report this to our support team to look into it in more detail: https://www.safe.com/support/

Thanks!

Hi @bknightos,

Sorry to hear you run into this issue!

Can you confirm that you are using the default example in step 4 of the doc you linked?

It looks like the password authentication is failing, but if you didn't change the default pw for the DB this is unexpected and not what we've seen in our tests.

 

 

If you didn't change any of the default settings please report this to our support team to look into it in more detail: https://www.safe.com/support/

Thanks!

@gerhardatsafe Thanks for your reply.

I followed the link and hadn't changed any passwords. That may happen in production however.

But for now I only need to get a system up and running.

The only difference to the linked example is that we are creating the cluster using terraform and installing helm within an azure release pipeline.

I will retry by creating the cluster manually and report to support team if I see the same issue.

Thanks.

Reply