Hi All,
One of the search head members in the search head cluster has a message: "Local KV Store has replication issues. See introspection data and mongod.log
for details. Local instance has state Recovering..".
What can I do to fix the issue?
When I checked with the kvstore status
command for this particular SH member, the status is shown as recovering.
On using the resync
command, even then the issue still exists.
Can you please let me know what steps should be followed to rectify the issue.
Will there be any impact on the performance of search heads?
Fixed the issue by following below steps:
1. Login to the SHM and navigate to the bin directory . Take a backup of KVstore using the below command
./splunk backup kvstore
Fixed the issue by following below steps:
1. Login to the SHM and navigate to the bin directory . Take a backup of KVstore using the below command
./splunk backup kvstore
Read this document: https://docs.splunk.com/Documentation/Splunk/9.0.3/Admin/ResyncKVstore
When you said "SHM" is it the Deployer, the Search Head Captain or the same SH with the issue?
The same message appears me today over one of the SHC members. To solve, I follow the recommendations over the member (I have 3 SH, the last one was the problematic).
First of all, you need to backup the KVstore. Splunk service should be running over the SH to backup.
Later, you need to restore the KVStore. Splunk service should be stopped before of the command execution.
All of these commands should be executed over the affected member. If you have multiple members affected, you would execute it on each affected member, or in all SH if all are being affected with this replication error.
Last Splunk version (8.2.4) recommends to move KVStore to a new technology. Currently, Splunk uses mmapv1 as Storage Engine, but it's recommended to move to wiredTiger. Follow the docs to migrate your KVStore to recommended engine: