QS Creating pool
Creating a Pool in the WebApp
Now the cluster is created the next step is to cluster a pool. This can either be an existing pool or one created using the webapp; if you already have pools configured skip ahead to the next section. In this example a mirrored pool will be created.
To create a pool, navigate to ZFS -> Pools and then click the + CREATE POOL
button on the main pools page to bring up the pool configuration page:
Fill out the Pool name field and select the desired structure of the pool from the Pool Mode list.
The cluster supports three types of layout when creating vdevs for a pool:
| Layout | Description |
|---|---|
mirror |
Each drive in the vdev will be mirrored to another drive in the same vdev. Vdevs can then be striped together to build up a mirrored pool. |
raidz1 |
One drive is used as a parity drive, meaning one drive can be lost in the vdev without impacting the pool. When striping raidz1 vdevs together each vdev can survive the loss of one of its members. |
raidz2 |
Two of the drives are used as parity drives, meaning up to two drives can be lost in the vdev without impacting the pool. When striping raidz2 vdevs together each vdev can survive the loss of two of its members. |
raidz3 |
Three of the drives are used as parity drives, meaning up to three drives can be lost in the vdev without impacting the pool. When striping raidz3 vdevs together each vdev can survive the loss of three of its members. |
jbod |
Creates a simple pool of striped disks with no redundancy. |
draid1 |
One drive is used as a parity drive, meaning one drive can be lost in the vdev without impacting the pool. When striping draid1 vdevs together each vdev can survive the loss of one of its members. |
draid2 |
Two of the drives are used as parity drives, meaning up to two drives can be lost in the vdev without impacting the pool. When striping draid2 vdevs together each vdev can survive the loss of two of its members. |
draid3 |
Three of the drives are used as parity drives, meaning up to three drives can be lost in the vdev without impacting the pool. When striping draid3 vdevs together each vdev can survive the loss of three of its members. |
dRAID (distributed raid) and RAIDZ are two different vdev layouts in ZFS, each offering distinct approaches to data redundancy and fault tolerance. RAIDZ is a traditional RAID-like structure that distributes data and parity information across multiple disks, while dRAID distributes hot spare space throughout the vdev, enabling faster rebuild times after drive failures.
Configure options according to your requirements - for a more in-depth discussion on options please see the HAC ZFS Tuning Guide:
Option |
Description |
|---|---|
Compression |
Compress data before it is written out to disk, choose either no compression, lz4 or zstd (on is an alias for lz4) |
Record Size |
The recordsize property gives the maximum size of a logical block in a ZFS dataset. Unlike many other file systems, ZFS has a variable record size, meaning files are stored as either a single block of varying sizes, or multiple blocks of recordsize blocks. |
Access Time |
Updated the access time of a file every time it is read or written. Recommended setting is off for better performance. |
Linux Access Time |
Hybrid setting meaning the access time is only updated if the mtime or ctime value changes or the access time has not been updated for 24 hours (on next file access). |
Alignment Shift |
Set to the sector size of the underlying disk - typically this is the value 12 for 4K drives (note some drives report a 512 byte sector size for backwards compatibility, but are in reality 4K; if unsure check manufacturers specifications) |
Extended Attributes |
This property defines how ZFS will handle Linux' eXtended ATTRibutes in a file system. The recommended setting is sa meaning the attributes are stored directly in the inodes, resulting in less IO requests when extended attributes are in use. For a file system with many small files this can have a significant performance improvement. |
Mirrored Pool
Mirrored pools are created by striping together individual mirrored vdevs.
Start by creating an individual mirrored DATA vdev (in this example a two way mirror is created,
but these could be three, four way mirrors etc). Select drives for the vdev from the
Available Disks list and click DATA to add them as data vdevs - in this example sdm and sdn
are used to create the first mirror:
To configure multiple striped mirrors, select the next
set of drives using the same number of drives as the existing data vdev, click DATA,
then from the popup menu select + New vdev (note that selecting vdev-0 would
extend the existing vdev rather than creating a new one):
This action will then add a further pair of mirrored drives to the pool layout, creating a mirrored stripe; additional drives are adding in the same manner:
Add further vdevs as required, here a log and a cache have been added:
Once configuration is complete, click SUBMIT and the pool will be created and displayed
in the main pools page ready for clustering. The configuration of the pool can be checked by clicking the
expand/collapse arrow on the left hand side of the pool entry:
