[AMBARI-17339] Ambari alert "NameNode Last Checkpoint" failing when NameNode is HA

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 2.4.0
• Assignee: Andrew Onischuk
• Reporter: Andrew Onischuk
• Created: 21/Jun/16 15:16
• Updated: 20/Oct/16 19:45
• Resolved: 22/Jun/16 10:03

Description

The "NameNode Last Checkpoint" alert is implemented in
alert_checkpoint_time.py. This script uses the value of **dfs.namenode.http-
address** property for querying the last checkpoint timestamp.

This works fine when the NN is not in HA mode. However when NN is in HA mode
this property is either not defined or might not be pointing to the right NN
instance.

The script needs to be improved such as to figure out if the NN is in HA mode.
If it's in HA mode do not use *dfs.namenode.http-address* property but
rather determine the correct URL to connect to similarly as how it's done in
the alert_ha_namenode_health.py

Issue Links

• duplicates: AMBARI-17107 Ambari Alert for NameNode Last Checkpoint in NameNode HA
• is duplicated by: AMBARI-17107 Ambari Alert for NameNode Last Checkpoint in NameNode HA
• links to: ReviewBoard