Plex is a logical copy of data in the NetApp ONTAP storage architecture and is part of an aggregate. A plex consists of one or more RAID groups and contains the aggregate’s data. A standard unmirrored aggregate uses one plex, while SyncMirror creates two plexes containing identical copies of the data.
A plex enables mirrored data storage across different sets of disks. If one of the mirrored plexes becomes unavailable due to a hardware failure, the system can continue operating using the other copy.
How a Plex Is Structured
In the NetApp architecture, the storage layers are logically organized as follows:
Disks → RAID Groups → Plex → Aggregate → Volumes
Each layer performs a specific function:
- Disks — physical or logically represented drives on which data is stored.
- RAID group — combines disks and provides protection against drive failures through redundancy.
- Plex — contains one or more RAID groups and represents one copy of the aggregate’s data.
- Aggregate — combines storage resources and provides space for volumes.
- Volume — a logical volume in which user data is stored.
Thus, a plex sits between RAID groups and the aggregate and is an element of ONTAP’s internal storage organization.
Plex in Unmirrored and Mirrored Aggregates
The number of plexes depends on the aggregate configuration.
An unmirrored aggregate uses one plex. It contains the RAID groups on which the data is stored.
A SyncMirror mirrored aggregate uses two plexes. They contain identical data but are located on different sets of disks:
- Plex 1 — one copy of the data
- Plex 2 — the mirrored copy
ONTAP synchronously maintains both copies in a consistent state. This architecture helps preserve access to data if one of the plexes becomes unavailable.
How Plex Is Related to SyncMirror
SyncMirror uses plexes as the basis for creating a mirrored aggregate. Instead of maintaining a single copy of the data, the system maintains two independent plexes.
During write operations, changes are written to both copies. Synchronous mirroring ensures that the data in both plexes remains consistent.
Separating plexes can provide protection against failures of individual infrastructure components. Depending on the configuration, the copies can be placed on different sets of disks and disk shelves or, in fault-tolerant architectures, on physically separated hardware.
Plex and RAID
A plex is not a type of RAID. These elements belong to different layers of the storage architecture.
A RAID group provides redundancy at the disk level. If an allowable number of drives fail, the data can be reconstructed using information stored on the remaining disks.
A plex combines one or more RAID groups and represents a copy of the aggregate’s data. Therefore, in a mirrored configuration, each plex has its own RAID groups.
This combination provides multiple layers of protection: RAID provides resilience against certain disk failures, while SyncMirror with two plexes maintains a second copy of the aggregate on a separate set of drives.
What Happens When a Plex Fails
If one plex in a mirrored aggregate becomes unavailable, the other continues to hold a copy of the data. Depending on the nature of the failure and the ONTAP configuration, data may remain available through the remaining plex.
After the cause of the failure has been resolved, the mirrored configuration must be restored. Data is resynchronized between the copies so that both plexes return to a consistent state.
However, having two plexes does not replace backup. If data is deleted or logically corrupted, the change may be replicated to both synchronized copies.
Plex vs. Volume: What Is the Difference?
Plex and volume belong to different layers of the NetApp architecture.
A plex is part of the internal organization of physical storage and contains RAID groups. A volume is a logical storage space for user data and is created using the resources of an aggregate.
Therefore, applications typically access data stored in volumes, while plexes are used by ONTAP to organize the underlying storage structure and mirroring.