Create snapshots of your TeamForge Orchestrate program files to restore system after a failure.
The goal of this topic is to help you take "snapshots" of your TeamForge Orchestrate installation.
Snapshots make a copy of all TeamForge Orchestrate program and configuration files and may be used to restore
your system in case of a failure. The snapshot process involves shutting down TeamForge Orchestrate services,
as such, we recommend that a single snapshot of each Orchestrate host be taken after installation and subsequent
upgrades, and periodic database backups be made to restore to a specific (more recent) data set.
See the Data back up and restore documentation to create periodic data backups
once you've created baseline snapshots.
Note: The TeamForge Orchestrate hostname cannot be changed when restoring from a snapshot. Please contact
support if you need to change the hostname of your TeamForge Orchestrate service.
Creating snapshots
A snapshot must be taken for each host running TeamForge Orchestrate services. If all of your TeamForge
Orchestrate services are on a single host, then only one snapshot is required. However, if you've spread
services across multiple hosts, then a single snapshot is required per host. The following steps describe
making a single snapshot, so you may need to repeat these steps if you have services spread across multiple
hosts.
-
Obtain orchestrate-snapshot-tools.tar.gz from the App server, located
in /opt/collabnet/orchestrate/tools/
-
Untar the orchestrate-snapshot-tools.tar.gz on the target host that you wish to snapshot.
-
Stop the TeamForge Orchestrate service(s) running on the target host.
See System Administration.
-
Run the take_snapshot.sh script with the appropriate argument.
The take_snapshot.sh script takes one of four arguments (mongodb, rabbitmq, orchestrate, all).
Run the script with the appropriate argument. For instance:
sudo ./take_snapshot.sh all (single host installation)
sudo ./take_snapshot.sh mongodb (multi-host, database server only)
sudo ./take_snapshot.sh rabbitmq (multi-host, message queue server only)
sudo ./take_snapshot.sh orchestrate (multi-host, application server only)
A snapshot will be created and this file should be moved to a safe location (off host).
The filename of a snapshot with options 'orchestrate' or 'all' will include the version number of
TeamForge Orchestrate.
-
Restart the TeamForge Orchestrate service(s). See System Administration.
Restore from a snapshot
-
Download the disconnected-media package corresponding to your version of Orchestrate. To
obtain a disconnected-media package, see Installation without internet access.
To restore from a snapshot, you must first download and install the disconnected media RPM package
matching your snapshot version number. For instance, if you have a snapshot file labeled 1.0.2, you should download
the disconnected media package version 1.0.2.
-
Install the disconnected-media RPM:
sudo rpm -ivh collabnet-orchestrate-disconnected-media-[Version].el6.x86_64.rpm
-
Run apply_snapshot.sh with the appropriate arguments
The apply_snapshot.sh script takes two arguments: the snapshot file you wish to
apply, and the type of snapshot you wish to apply (mongodb, rabbitmq, orchestrate, all).
Run the script with the appropriate argument. For instance:
sudo ./apply_snapshot.sh snapshot_file.tar.gz all (single host installation)
sudo ./apply_snapshot.sh snapshot_file.tar.gz mongodb (multi-host, database server only)
sudo ./apply_snapshot.sh snapshot_file.tar.gz rabbitmq (multi-host, message queue server only)
sudo ./apply_snapshot.sh snapshot_file.tar.gz orchestrate (multi-host, application server only)
-
Restore desired data backup. See Data back up and restore.
-
Restart the TeamForge Orchestrate service(s). See System Administration.