A customer is experiencing 'Search peer [peer-name] is not responding' errors during indexer maintenance. Which configuration parameter in server.conf should be tuned to prevent search failures during rolling restarts?
Trap 1: peer_down_timeout
peer_down_timeout governs how long a search head waits before declaring a peer down; raising it delays the failure rather than preventing it during rolling restarts. It tempts because it directly references peer availability, and would be right when transient network latency, not maintenance restarts, causes peers to drop.
Trap 2: max_search_time
This limits individual search job duration.
Trap 3: heartbeat_timeout
This controls communication between indexers and the master node.
- A
peer_down_timeout
Why it fails: peer_down_timeout governs how long a search head waits before declaring a peer down; raising it delays the failure rather than preventing it during rolling restarts. It tempts because it directly references peer availability, and would be right when transient network latency, not maintenance restarts, causes peers to drop.
- B
max_search_time
Why it fails: This limits individual search job duration.
- C
restart_timeout
This allows for longer wait times during planned maintenance.
- D
heartbeat_timeout
Why it fails: This controls communication between indexers and the master node.
- E
search_retry_timeout
Why it fails: This is not a valid parameter.