Information required for debugging issues on the Tendrl stack

In the short-to-medium term we’ll be adding a log collector script/utility which can be run on each node of the cluster and package the logs in a structured pay-load. This document intends to build up the set of data points which will be addressed by the script/utility.

Specific to issues (failure/errors/incomplete) observed during import

General information

Important

This information is required for each of the cluster nodes.

• output of ps aufx

• output of netstat -ntlp

• output of rpm -qa | grep tendrl

• the contents of the /etc/hosts file

• output of hostname

• output of ip addr ls

Additionally, the output of the following from the python 2.7.x console:

import socket
socket.getfqdn()
socket.gethostname()
socket.gethostbyaddr([ip]) # where [ip] is the IP address, run this once for each IP of the system

Gluster specific issues

Important

This information is required from each of the cluster nodes, in addition to the general information specified above.

If the import issues are observed on a Gluster cluster, we need:

• Output of gluster peer list