Hostd goes unresponsive when using hpsa driver version prior to 2.0.24

Products

• VMware

Issue/Introduction

Symptoms:

• Hostd goes unresponsive
• This is also encountered in vSAN environment.
• Restarting the hostd service may bring the service back online.
• In /var/log/hostd.log you see entries :

2018-12-05T23:45:02.257Z cpu0:697159)ALERT: hostd detected to be non-responsive
2018-12-06T00:20:01.897Z cpu44:713781)ALERT: hostd detected to be non-responsive
2018-12-06T00:55:01.553Z cpu34:731608)ALERT: hostd detected to be non-responsive
2018-12-06T01:25:02.069Z cpu10:745863)ALERT: hostd detected to be non-responsive

•    Check the current version of hpsa driver  and plugin installed.

​​​​​​​For example​​​​​​

#esxcli software vib list | grep -i hpsa
nhpsa                          2.0.22-3vmw.650.2.50.8294253         VMW     VMwareCertified   2018-12-04
lsu-hp-hpsa-plugin             2.0.0-6vmw.650.2.50.8294253          VMware  VMwareCertified   2018-12-04

Note:The preceding log excerpts are only examples.Date,time and environmental variables may vary depending on your environment.

Cause

hpsa plugin depends on the ssacli to fetch device information, which will use command passthrough to communicate with the driver. If ssacli blocked in the driver, it can lead to all the threads being blocked  and so the hostd will go to  non-responsive state.

Resolution

To resolve this issue upgrade the hpsa driver version to 2.0.24-1OEM from Customer Connect Downloads

Workaround:
To workaround this issue, follow the below workarounds

1. Stop hostd service:
# /etc/init.d/hostd stop

2. Remove the lsu hpsa plugin:
# localcli software vib remove -n lsu-hp-hpsa-plugin

3. Restart hostd service:
# /etc/init.d/hostd start

Additional Information

Impact/Risks:
Removing the plugin will cause the functionalities of the plugin to be lost.