File management appliance (FMA) upgrade

Step 1: Verify that recalls are working in FMA-HA environments

1.Stop the recall daemons for your type of archiving on the FMA:
celerracallback stop
fpolicycallback stop

2.After stopping each service, recall a few archived files to verify that access to archived data is uninterrupted. This will verify that the recalls are coming through the FMA-HA.


3.Start the services on the FMA
celerracallback start
fpolicycallback start

4.Stop the recall daemons for your type of archiving on the FMA-HA.

5.After stopping each service, recall a few archived files to verify that access to archived data is uninterrupted. This will verify that the recalls are coming through the FMA.

6.Start the services on the FMA-HA
celerracallback start
fpolicycallback start

Step 2:  Verify that log rotation is working properly

#/usr/sbin/logrotate -v /etc/logrotate.conf

Review the output and check the specified log files to determine if the log files are rotated. Any error that occurs during the upgrade process is written to a log. If log rotation is not functioning properly, logs used to troubleshoot upgrade problems might be overwritten.

Step 3: Verify that the FMA database is functioning properly

#service postgresql status

Step 4: Vacuum DB

# rffm doDBMaintenance

Step 5:  Backup database and FMA configuration

1. From the FMA CLI, enter a screen session by typing:

#screen
2. Stop the filemanagement service by typing:

#filemanagement stop
3. Create a backup of the FMA by typing:

#fmbackup

The process writes a backup file to
/var/fmbackup.<machine_name>.<timestamp>.tgz.
Copy the fmbackup file to another system. If needed for disaster recovery, restore the backup with the command:
fmrestore /var/fmbackup.<machine_name>.<timestamp>.tgz

4. Once the backup file is successfully created, stop the postgresql service by typing:

#service postgresql stop

Step6: UPG Upgrade

1. Upload the .UPG file to the FMA using WinSCP.

2. Copy the rfupgrade script that was downloaded: Obtain the CTA Software Image and rfupgrade to /opt/rainfinity/filemanagement/bin

3. Confirm that this is the correct rfupgrade script by typing:
# cd /opt/rainfinity/filemanagement/bin
# mv rfupgrade rfupgrade.old
# unzip rfuprade.zip
# md5sum rfupgrade
# md5sum /tmp/rf_7.5-220.i686.upg

4. Stop the filemanagement service if it is currently running.
# filemanagement stop

5. Start the upgrade.

/opt/rainfinity/filemanagement/bin/rfupgrade /tmp/rf_7.5-220.i686.upg

Note: The upgrade process begins with a database pretest script that checks to see if the FMA databases are consistent between the old and new releases. If the pretest finds inconsistencies, the upgrade will exit with the error message: Failed to upgrade database. Contact EMC Customer Support to correct the problem before restarting the upgrade.
If no problems are encountered, the process upgrades the rpm files.
To exit the screen session once the database upgrade has started, press Ctrl - a Ctrl ñ d.

6. Once the upgrade concludes, check if the callback daemons have started by typing:
celerracallback status
fpolicycallback status
If they are not running start the callback daemons by typing:
celerracallback start
fpolicycallback start

Step6: Validate

rffm version