Not able to access Gluu Cluster Manager if primary server goes down

By: Heero Punjabi Account Admin 21 Nov 2018 at 7:43 a.m. CST

11 Responses
Heero Punjabi gravatar
If the Primary Server goes down, I am not able to access the Gluu Cluster Manager, But load balancer and remaining nodes working fine. How can i resolve this issue ?
Ubuntu 16.0
closed

Answers

By Mohib Zico Account Admin 21 Nov 2018 at 8:01 a.m. CST

Copy
Mohib Zico gravatar
If Primary Gluu Server goes down, you can't access Cluster Manager GUI with browser?

By Heero Punjabi Account Admin 21 Nov 2018 at 11:38 p.m. CST

Copy
Heero Punjabi gravatar
I stop the primary server, sometime the cluster manager not able to access. there is no network issues, but sometime I got access to the web ui for cluster manager, but showing something like this screenshot. not able to manage the remaining gluu nodes. So if the primary server goes down, how is the slave gluu cluster will work ?, one of the slaves promote to primary server ?
Screen_Shot_2018-11-22_at_10.59.58_AM.png

By Chris Blanton user 21 Nov 2018 at 11:43 p.m. CST

Copy
Chris Blanton gravatar
Heero, i believe we fixed this in a later version. On the cluster manager machine you need to upgrade with `pip install -U clustermgr` with the user you installed Cluster Manager with. Then run `clustermgr-cli restart` and you should be running the latest version.

By Heero Punjabi Account Admin 22 Nov 2018 at 3:41 a.m. CST

Copy
Heero Punjabi gravatar
Thanks for your answer.. Please respond to these as well, 1) So if the primary server goes down, how is the slave gluu cluster will work ?, one of the slaves promote to primary server ? 2) Is the cluster working as master - slave configuration ?

By Mohib Zico Account Admin 22 Nov 2018 at 4:03 a.m. CST

Copy
Mohib Zico gravatar
I think it's master-master setup. Chris will confirm though.

By Heero Punjabi Account Admin 22 Nov 2018 at 8:11 a.m. CST

Copy
Heero Punjabi gravatar
Hi chris, please find the attached snapshot, if the primary server goes down, how is the slave gluu cluster will work ?, the gluu authentication web ui showing this error.
gluu-error.png

By Chris Blanton user 22 Nov 2018 at 2:29 p.m. CST

Copy
Chris Blanton gravatar
Heero, > 2) Is the cluster working as master - slave configuration ? Gluu Server clusters are entirely master - master, so there are no slaves. You generally get full redundancy when configured correctly. > if the primary server goes down, how is the slave gluu cluster will work ?, the gluu authentication web ui showing this error. Can you tell me how your cluster is configured? Are you using NGINX as a load balancer or something else? Redis Cache or LDAP Cache? Also send me your logs in `/opt/gluu/jetty/identity/logs/`

By Chris Blanton user 22 Nov 2018 at 2:33 p.m. CST

Copy
Chris Blanton gravatar
> 1) So if the primary server goes down, how is the slave gluu cluster will work ?, one of the slaves promote to primary server ? I'm not sure if you mean Gluu Server, or Cluster Manager, but `primary` is only so Cluster Manager knows which server to use as a seed for the other servers for their initial configuration. After that point, Cluster Manager simply monitors services and is used for operational tasks related to ease of maintenance of the Cluster.

By Heero Punjabi Account Admin 23 Nov 2018 at 2:53 a.m. CST

Copy
Heero Punjabi gravatar
**Can you tell me how your cluster is configured? Are you using NGINX as a load balancer or something else? Redis Cache or LDAP Cache? Also send me your logs in /opt/gluu/jetty/identity/logs/** I configured with nginx and redis cache, but the error occurring only when i am removing the old primary server from cluster manager, log from opt/gluu/jetty/identity/logs/2018_11_23.jetty.log : **Caused by: org.xdi.exception.ConfigurationException: Failed to create LDAP connection pool!** Please find the screenshot as well
gluu-error-2.png

By Mohib Zico Account Admin 29 Nov 2018 at 7:17 a.m. CST

Copy
Mohib Zico gravatar
Hi Heero, >> I configured with nginx and redis cache, but the error occurring only when i am removing the old primary server from cluster manager, So you are saying that if your remove 'seed Gluu Server' ( which is the primary server of cluster ) from cluster pool; 2nd node going down?

By Mohib Zico Account Admin 07 Dec 2018 at 1:57 a.m. CST

Copy
Mohib Zico gravatar
Hello, Closing this ticket due to inactivity. Please reopen when required.

Post an answer