<

Reverse path filtering setting between RHEL 5 vs RHEL 7.

Published on
3,584 Points
584 Views
Last Modified:
This is about the change in the value setting for the rp_filter parameter in Linux.

The Background


Recently, we upgraded our Oracle database server Hardware and OS.
Our Oracle database was on a 2-node 12g RAC.
We went from RHEL 5.9 to 7.1

The plan is as below


  •  DBA will delete the node on the old hardware (Oracle support suggested this).
  •  Storage Admin will un-present the storage from the old hardware.
  •  Sys Admin will prepare the new hardware and give it to Storage Admins.
  •  Storage Admin will represent the storage of the new hardware.
  •  DBA will re-add the node using addnode.sh.


The Issue I faced


All was fine until removal part.

 
But, while trying to add the node, I had the trouble at the cluster verify stage, while running clufy with -pre-flag as it sensed the bondings were not working.


When researched, finally the network team found out and told the Sys Admin that there is a change in the behavior of the setting for the parameter net.ipv4.conf.default.rp_filter.


Summary of the finding


RHEL7, RHEL6, and RHEL5 ship with a default /etc/sysctl.conf that sets this sysctl to 1, but the meaning of this value is different between the RHEL6 and the RHEL5 kernel.


So once this is set to 0 in RHEL 7, our network between the old hardware and new hardware could communicate with each other and the addnode.sh could successfully copy the Cluster Software from the surviving node to the newly added server and made it part of the Oracle cluster.


I hope you found this helpful.



0
Ask questions about what you read
If you have a question about something within an article, you can receive help directly from the article author. Experts Exchange article authors are available to answer questions and further the discussion.
Get 7 days free