Redis probe can not connect to redis node in a cluster (reason: Connection and/or inventory update failure)

Products

• DX Unified Infrastructure Management (Nimsoft / UIM)

Issue/Introduction

When configuring the redis probe for a clustered Redis server, the nodes are not seen in the probe UI and this alarm is seen in OC:

NFM Redis Server is not responding (reason: Connection and/or inventory update failure) (Unable to connect to to zz.zz.zz.zz:56379)

where zz.zz.zz.zz is the Redis main node (master)

Also, redis logs show the following:

[Data Collector - NFM Redis Server, redis] Getting Cluster Metrics
[Data Collector - NFM Redis Server, redis] Scheduling executor service to collector metrics
[pool-896-thread-5, redis] Exception while executing commands : Unable to connect to xx.xx.xx.xx:6379
[pool-896-thread-4, redis] Exception while executing commands : Unable to connect to xx.xx.xx.yy:6379

where xx.xx.xx.xx and xx.xx.xx.yy are the internal IP addresses of the Redis nodes (slaves) in the cluster.

Environment

UIM version : ANY

redis probe version : Redis v1.01

Redis version : Any version supported

Cause

If you monitor a clustered Redis environment, you need to provide the IP address or hostname of any cluster node you monitor. If the robot where the probe is deployed do not have access to the Redis nodes/slaves, the connection will fail.

 

Resolution

- Expose the internal IP of the Redis node and configure the redis probe to connect to this node using the external IP address,

or

- deploy a UIM robot on a host that does have access to the Redis nodes and deploy there the redis probe.