Shared storage is required for files uploaded to Orchestrator, as well as for the associated
MySQL data files and ACM files. The shared storage must meet the following criteria:
- High availability (for example, with RAID arrays) to avoid a single point of
failure
- Completely reliable and accessible 100% of the time, so ACM can secure highly available
Orchestrator operations and prevent data corruption.
Minimum configuration requirements are the following:
- MySQL data: 20GB
- ACM: 1GB
- Orchestrator var data: 20GB
Because ACM is not migrating the mount points over failover, MySQL data, Orchestrator
var data, and ACM data must be mounted and visible on both nodes
simultaneously.
Shared File Systems Compatible with ACM
Aspera recommends dedicating storage for ACM whenever possible in order to avoid creating
I/O bottlenecks when large packages are being transferred to shared storage at very high
speeds.
ACM supports SAN and NAS external storage. However, be aware of the following
limitations:
- A SAN storage requires a cluster file system for the Aspera high availability
environment to function properly. Block-level storage access will not work; only
file-level storage access is suitable. Files must be visible from both nodes concurrently.
If a file is changed on one node, the other node must be able to immediately incorporate
the changes.
- A storage solution with NFS Version 3 may cause problems, because NFS3 does not
automatically release the locks held by a client. This is critical for MySQL, where only
one node at a time may write to the shared disk. With NFS3, if the active node that is
running MySQL goes down, the passive node is unlikely to be able to start MySQL when it
takes over, because some MySQL files may still be locked on the shared storage. With NFS4
technology, all locks associated with a TCP session that times out will be automatically
released, making MySQL automatic failover possible.
- The NFS mount option noac must be used with the /mnt/shared/orchestrator/acm_files/
mount point for ACM to function properly.
Important: Using noac on any
mount point other than /mnt/shared/orchestrator/acm_files/ may lead to severe
performance issues.
See Mounting Remote File Systems on Each Orchestrator Server for more information.
- CIFS is not supported.
Folders that Must be Shared
For
Orchestrator on-premise high availability installation, the folders and other data sources
listed below must be set up with shared external storage and shared between the nodes.
- The run
folder:
/opt/aspera/var/run/orchestrator/
Important run-time
information is stored in this folder. For example, the file trigger history is stored
here, and thus each node has the same perspective of already-triggered files.
Alternatively, the history may also be kept in the shared MySQL data. In orchestrator.yml,
set the parameter persistence_mode to either drive or db (both options are
correct values).
- The config
folder:
/opt/aspera/var/config/orchestrator/
The plugin directories
are present in this folder (in the actions subdirectory). Importing a plugin is done on
only one node.
- The archive
folder:
/opt/aspera/var/archive/orchestrator/
The plugins are
archived in this folder; sharing it allows the nodes to share the same view of archived
plugins.
- MySQL data from the following folder:
/opt/aspera/common/mysql/data
- ACM data.
On servers that have other Aspera products installed, do not share the following
folder:
/opt/aspera/var/
Instead, share the these
folders:
/opt/aspera/var/config/orchestrator
/opt/aspera/var/run/orchestrator
/opt/aspera/var/archive/orchestrator