Exadata Deployment Error : CELL-02559

While working on deploying Exadata for one of my client, I got following errors during deployment step # 2. Error clearly states that there is communication error between MS and CELLSRV. In my case this error was somewhat miss leading. After investigating for few hours, I found that ( /opt/oracle/cell/cellsrv/deploy/config/cellinit.ora ) files on Exadata storage cells contain wrong IP addresses. You need to update cellinit.ora files with correct ip addresses and restart all cell services to resolve this error.  You can use following steps to verify and resolve this issue.

[root@XXXX-db01linux-x64]# ./install.sh -cf client.xml -s 2 
 Initializing 
 Executing Update Nodes for Eighth Rack 
 running setup on: XXXX-db02 
 Completed copying resourcecontrol file to XXXX-cel02-adm 
 running setup on: cvdp2-cel02-adm 
 XXX-cel02-adm needs total CPU cores set from 20 to 10 
 ERROR: 
 CELL-02559: There is a communication error between MS and CELLSRV. 
 Error: Errors occured... 
 Collecting diagnostics... 
 ERROR: 

  1. Verify cell services using (list cell detail) on all nodes.

Note : – As you can see cellsrv service is stopped and causing original error. Also make a note of IP addresses, we need to verify them with cellinit.ora file and private InfiniBand ip addresses.

  1. Restart all cell services

You can try re-starting all services and see you get any errors

  1. Check InfiniBand IP addresses using ifconfig command

  1. Check contents of cellinit.ora

Note : – Make sure these are correct InfiniBand private IP addresses. In my case they were wrong.

  1. Update cellinit.ora with correct InfiniBand IP addresses

Make sure to backup cellinit.ora file and Edit cellinit.ora file with correct IP addresses using vi editor.

  1. Restart all cell services

Now you can restart all services and hopefully , you will not run into any issues this time.

alter cell restart services all