Home > Storage > PowerScale (Isilon) > Product Documentation > Management and Migration > Dell PowerScale: Non-Disruptive Upgrade Best Practices > OneFS upgrade
This section will focus on the guideline and methodologies of troubleshooting OneFS upgrade issues.
To monitor the high-level OneFS upgrade status use the following command
isi upgrade view –interactive
Starting from OneFS 8.2.0, it is possible to detect upgrade hangs. If a OneFS upgrade is not making any progress after 15 minutes, the CELOG sends a notification. In general, CELOG Alerts will be generated for three causes of upgrade hangs:
For these three reasons, there are three newly introduced CELOG events:
400150007: Upgrade Hang – unable to communicate with Upgrade Agent on devids: {devids}
400150006: Upgrade Hang – Upgrade Agent unable to make progress on devids: {devids}
400150008: Upgrade Hook Hang - {hook} on devids: {devids}
Use the following command to check the event list for unresolved Upgrade Hang events:
isi event events list
The location of the OneFS upgrade logs varies between MRs. They can be categorized into two classes:
Starting from OneFS 8.2.0, isi_upgrade_logs supports to filter and extract for specific upgrade logging information. The following list shows the supported filters. The filters can be used in combination.
--guid - dump the logs for the node with the supplied guid
--devid - dump the logs for the node/s with the supplied devid/s
--lnn - dump the logs for the node/s with the supplied lnn/s
--process - dump the logs for the node with the supplied process name
--level - dump the logs for the supplied level range
--time - dump the logs for the supplied time range
--metadata - dump the logs matching the supplied regex
An example is shown below to extract all the upgrade log generated by the process isi_upgrade_agent_d on PowerScale node 3 and 4:
isi_upgrade_logs --lnn=3,4 --process=/usr/bin/isi_upgrade_agent_d
Note: When upgrading into 8.1.1.0 or later release from a pre 8.1.1.0 release, the logging will follow the old format.
In case you encounter some errors during the OneFS upgrade process, it is recommended to investigate the latest upgrade log first. After you determine the root cause and get the issue fixed, you can use the following command to retry the last upgrade action.
tme-sandbox-2# isi upgrade retry-last-action all
If it does not work, use the following command to roll back the upgrade:
tme-sandbox-2# isi upgrade rollback
Note: Rollback can lead to a simultaneous reboot process of all the nodes in the PowerScale cluster. Be careful to initiate OneFS rollback.