Troubleshooting
Problem
nzhw -issues reports standby host issues nzhw -issues Description HW ID Location Role State Security ----------- ----- ---------------- ------ ----------- -------- Host 1009 rack1.host2 Spare Unreachable N/A Fan 1055 rack1.host2.fan1 Active Unreachable N/A Fan 1056 rack1.host2.fan2 Active Unreachable N/A Fan 1057 rack1.host2.fan3 Active Unreachable N/A PowerSupply 1058 rack1.host2.pwr1 Active Unreachable N/A PowerSupply 1059 rack1.host2.pwr2 Active Unreachable N/A
Symptom
nzhw -issues reports standby host unreachable. passwordless ssh does not work from active host to standby host as root user
Cause
The NPS system makes use of being able to connect to the standby host over ssh without needing a password to perform various activities dealing with cluster operation and checking of health.
If for any reason the system is unable to use ssh to the standby host without needing a password the operation it is trying to do will fail.
When the nzhealthcheck script fails to be able to connect because passwordless ssh is broken it will report the standby host as unreachable and the components it monitors as unreachable.
nzhw -issues reports standby host issues
nzhw -issues
Description HW ID Location Role State Security
----------- ----- ---------------- ------ ----------- --------
Host 1009 rack1.host2 Spare Unreachable N/A
Fan 1055 rack1.host2.fan1 Active Unreachable N/A
Fan 1056 rack1.host2.fan2 Active Unreachable N/A
Fan 1057 rack1.host2.fan3 Active Unreachable N/A
PowerSupply 1058 rack1.host2.pwr1 Active Unreachable N/A
PowerSupply 1059 rack1.host2.pwr2 Active Unreachable N/A
Environment
NPS PureData System for Analytics
Diagnosing The Problem
Test to make sure the root user is able to ssh between both hosts without needing a password.
From HA1 # ssh ha2 and then from HA2 # ssh ha1. If either command prompts for a password , passwordless ssh is broken.
Resolving The Problem
This may be caused by many issues with ssh.
Missing or corrupted key files /root/.ssh/id_rsa or /root/.ssh/id_rsa.pub
Incorrect permissions or ownership of the directories or key files.
Mis-configuration of /etc/ssh/sshd_config
Mis-configuration of /root/.ssh/authorized_keys
sshd service not running
Use ssh in verbose mode to check for issues.
# ssh -vvv ha2 Note that the more -v's the more detailed the information. Start with one -v and raise as needed.
ssh -v will identify permission issues fairly quickly.
/etc/ssh/sshd_config should have either
PermitRootLogin without-password
or
PermitRootLogin yes (this is the default if the option is commented out)
If changes are made to /etc/ssh/sshd_config then the sshd daemon will need to be restarted.
# service sshd restart
Please note that there is no relationship between the hosts root passwords and the ability to passwordless ssh between them.
Was this topic helpful?
Document Information
Modified date:
17 October 2019
UID
swg21974415