In this guide, we are going to learn how to install and setup GlusterFS storage cluster on CentOS 8. GlusterFS is an opensource distributed and scalable network file system that clusters various disk storage resources into a single global namespace. It is suitable for data-intensive tasks such as cloud storage and media streaming. Some of the common features for GlusterFS include;
- Can scale to several petabytes thus can handle thousands of clients.
- Provides high availability through data mirroring. It also supports self healing mechanism that restores data to the correct state following recovery with nearly no overhead.
- Provides a unified global namespace that clusters disk and memory resources into a single pool that ensures load balanced I/O.
- Uses elastic hash algorithm to locate data in the storage pool hence linear performance scaling.
- Provides elastic volume manager in which data is stored in logical volumes that are abstracted from the hardware and logically partitioned from each other. This ensures that storage can be added or removed while data continues to be online with no application interruption.
- Gluster is fully POSIX-compliant and does not require any unique APIs for data access.
- Supports industry standard protocols like NFS, SMB, CIFS, HTTP and FTP.
Install and Setup GlusterFS Storage Cluster on CentOS 8
Ensure that you have at least three storage nodes. Check our deployment architecture below.
Also, ensure that you have an extract virtual disk apart from the root disk where the OS is installed. In our case above, we have an extra disk,
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT sda 8:0 0 8G 0 disk ├─sda1 8:1 0 1G 0 part /boot └─sda2 8:2 0 7G 0 part ├─cl-root 253:0 0 6.2G 0 lvm / └─cl-swap 253:1 0 820M 0 lvm [SWAP] sdb 8:16 0 2G 0 disk └─sdb1 8:17 0 2G 0 part sr0 11:0 1 1024M 0 rom
If using hostnames, ensure that they are resolvable.
Ensure that time is tightly synchronized between the three nodes. If you need to setup your NTP server on a CentOS 8 system, simply follow the link below;
Partition and Format the GlusterFS Storage Drives
On each GlusterFS node, format and create a file system on the storage drive. You can use any partitioning method.
Format and create a primary partition on the disk.
parted /dev/sdb mklabel msdos
parted -a opt /dev/sdb mkpart primary xfs 0% 100%
Create a Filesystem on the disk. XFS is used here.
mkfs.xfs /dev/sdb1 -L gfsdisk
Mount GlusterFS Storage Drives (Bricks)
On each node, create a mount point for the GlusterFS storage drive, here in called bricks. A brick is a basic unit of GlusterFS storage.
mkdir -p /export/gfsbrick
Mount the GlusterFS storage drive on the bricks directory created above.
mount /dev/sdb1 /export/gfsbrick/
To automount the drive on system boot, simply update the /etc/fstab configuration file as follows;
echo "/dev/sdb1 /export/gfsbrick/ xfs defaults 1 2" >> /etc/fstab
Check the mounting;
df -hTP /dev/sdb1
Filesystem Type Size Used Avail Use% Mounted on /dev/sdb1 xfs 2.0G 47M 2.0G 3% /export/gfsbrick
Install GlusterFS 7 on CentOS 8
Install GlusterFS Repo
Now that the prerequisites part is done, proceed to install GlusterFS on all nodes. GlusterFS is not available on the default repos. Hence, run the command below to create the repo.
wget -P /etc/yum.repos.d/ https://download.gluster.org/pub/gluster/glusterfs/7/LATEST/CentOS/glusterfs-rhel8.repo
Enable PowerTools repos in order to be able to install other required GlusterFS dependencies.
dnf config-manager --set-enabled PowerTools
Install GlusterFS server package on CentOS 8 by running the command below;
dnf install glusterfs-server
Running GlusterFS on CentOS 8
Start and enable GlusterFS daemon to run on system boot by executing the command below;
systemctl enable --now glusterd
To check the status of GlusterFS daemon;
systemctl status glusterd
● glusterd.service - GlusterFS, a clustered file-system server Loaded: loaded (/usr/lib/systemd/system/glusterd.service; enabled; vendor preset: disabled) Active: active (running) since Fri 2020-03-20 19:51:37 EAT; 12s ago Docs: man:glusterd(8) Process: 2739 ExecStart=/usr/sbin/glusterd -p /var/run/glusterd.pid --log-level $LOG_LEVEL $GLUSTERD_OPTIONS (code=exited, status=0/SUCCESS) Main PID: 2740 (glusterd) Tasks: 9 (limit: 5047) Memory: 3.9M CGroup: /system.slice/glusterd.service └─2740 /usr/sbin/glusterd -p /var/run/glusterd.pid --log-level INFO Mar 20 19:51:37 gfs01.kifarunix-demo.com systemd: Starting GlusterFS, a clustered file-system server... Mar 20 19:51:37 gfs01.kifarunix-demo.com systemd: Started GlusterFS, a clustered file-system server.
Allow GlusterFS on Firewall
In order for the GlusterFS nodes to be able to communicate, you need to open certain GlusterFS ports or services on a firewall. The
24007-24008/TCP are used for the communication between nodes, while
24009-24108/TCP are required for client communication. To make this easy, simply use the service,
firewall-cmd --add-service=glusterfs --permanent
Configure GlusterFS Trusted Storage Pool
To configure GlusterFS storage cluster, you first need to create a
trusted storage pool (TSP) which basically is a network of storage servers. TSP can be created by adding storage nodes to the TSP using the
gluster peer probe command as shown below.
To probe other nodes, eg gfs02 and gfs03 from gfs01 in our case, run the command below. Replace the names of the nodes accordingly. You can probe other nodes from any node.
gluster peer probe gfs02
gluster peer probe gfs03
To verify cluster peering status, run the command below from any node. Eg from node 1, gfs01;
gluster peer status
Number of Peers: 2 Hostname: gfs01.kifarunix-demo.com Uuid: b2003be4-4bc9-4125-a2a1-0d46e4b2476f State: Peer in Cluster (Connected) Hostname: gfs02 Uuid: 6bc3b730-b7bb-4d24-865a-c33a3db4489d State: Peer in Cluster (Connected)
To list all nodes in a TSP, eg from GlusterFS node 02, gfs01;
gluster pool list
UUID Hostname State 6bc3b730-b7bb-4d24-865a-c33a3db4489d gfs02 Connected 26cdb6a4-ad15-41e7-bb72-2da0a9906fa6 gfs03 Connected b2003be4-4bc9-4125-a2a1-0d46e4b2476f localhost Connected
You can be able to add more serves or remove servers from the GlusterFS storage pool;
For example, to add a new server, you need to probe it from a node already in the pool. Replace the server-name accordingly.
gluster peer probe server-name
To detach a server from the pool;
gluster peer detach server-name
Setup GlusterFS Storage Volume
A GlusterFS storage volume can be created from the bricks that were created above. In this case, we have created bricks mounted on
/export/gfsbrick on each node.
GlusterFS Storage Volume Types
There are different configuration settings to consider while creating GlusterFS storage volume depending on various operational needs.
Distributed: Files are distributed across the bricks in the volume.
Replicated: Files are replicated across the bricks in the volume. It ensures high storage availability and reliability.
Distributed Replicated: Files are distributed across the replicated bricks in the volume. Ensures high-reliability, scalability and improved read performance.
Arbitrated Replicated: Files are replicated across two bricks in a replica set and only the metadata is replicated to the third brick. Ensures data consistency.
Dispersed: Files are dispersed across the bricks in the volume.
Distributed Dispersed: Data is distributed across the dispersed sub-volume.
The GlusterFS volume can be created using the command;
gluster volume create <NEW-VOLNAME> [stripe <COUNT>] [[replica <COUNT> [arbiter <COUNT>]]|[replica 2 thin-arbiter 1]] [disperse [<COUNT>]] [redundancy <COUNT>] [transport <tcp|rdma|tcp,rdma>] <NEW-BRICK> ...<TA-BRICK>
man gluster for more details.
Setup Replicated GlusterFS Storage Volume
In this demo, we are going to learn how to setup replicated glusterfs storage volume.
Create a mount point for the volume on the brick on each cluster node. In our case, our brick is mounted on
Next, created a replicated storage volume. For example, to create a replicated GlusterFS storage volume with three nodes called
gluster volume create gfsvol01 replica 3 transport tcp gfs01:/export/gfsbrick/gfsvol01 gfs02:/export/gfsbrick/gfsvol01 gfs03:/export/gfsbrick/gfsvol01
You can run the command from any node, replacing the names of the other peers and the associated bricks accordingly.
If everything goes well, you should get an output like;
volume create: gfsvol01: success: please start the volume to access data
After that, start the GlusterFS volume. Replace
gfsvol01 with the name of your volume.
gluster volume start gfsvol01
Verify GlusterFS Volumes
Once you have created and started the volumes, you can verify by running the command below from any node in the cluster.
gluster volume info all
Volume Name: gfsvol01 Type: Replicate Volume ID: 9361067e-da82-411c-9c83-2ab1c2b888aa Status: Started Snapshot Count: 0 Number of Bricks: 1 x 3 = 3 Transport-type: tcp Bricks: Brick1: gfs01:/export/gfsbrick/gfsvol01 Brick2: gfs02:/export/gfsbrick/gfsvol01 Brick3: gfs03:/export/gfsbrick/gfsvol01 Options Reconfigured: transport.address-family: inet storage.fips-mode-rchecksum: on nfs.disable: on performance.client-io-threads: off
There you go. You have successfully installed and setup replicated GlusterFS storage cluster.
Want to set a different type of GlusterFS volume? Consult GlusterFS documentation on setting up volume types.
If anything, you can always check the log file,
/var/log/glusterfs/glusterd.log for any would be issues.
Install and Setup GlusterFS client on CentOS 8
Now that you have a GlusterFS storage cluster running, you can setup a client to be able to utilize this storage. Gluster volumes can be accessed using the Gluster Native Client or via NFS v3.
In this demo, we are going to learn how to use the native GlusterFS client on CentOS 8.
Thus, create the GlusterFS repositories on CentOS 8;
wget -P /etc/yum.repos.d/ https://download.gluster.org/pub/gluster/glusterfs/7/LATEST/CentOS/glusterfs-rhel8.repo
Run system update;
Install Gluster Native Client on CentOS 8
dnf install glusterfs glusterfs-fuse
Mounting Gluster Storage Volumes on Client
To manually mount the Gluster storage volumes on a client, use the command;
mount -t glusterfs HOSTNAME-OR-IPADDRESS:/VOLNAME MOUNTDIR
t glusterfsspecifies the GlusterFS filesystem.
HOSTNAME-OR-IPADDRESSis any gluster node eg, gfs01, gf02 or gfs03. Ensure hostnames are resolvable.
/VOLNAMEis the name of the GlusterFS storage volume e.g
gfsvol01in this demo.
MOUNTDIRis the mount point on the client.
Thus, to mount the Gluster storage volume from glusterfs node 01, gfs01, our mount command would be like;
mount -t glusterfs gfs01:/gfsvol01 /media/gfs/
Confirm the mounting;
Filesystem Type Size Used Avail Use% Mounted on devtmpfs devtmpfs 1.4G 0 1.4G 0% /dev tmpfs tmpfs 1.4G 0 1.4G 0% /dev/shm tmpfs tmpfs 1.4G 9.9M 1.4G 1% /run tmpfs tmpfs 1.4G 0 1.4G 0% /sys/fs/cgroup /dev/mapper/cl-root xfs 8.0G 5.7G 2.4G 71% / /dev/sda1 ext4 976M 244M 666M 27% /boot ... gfs01:/gfsvol01 fuse.glusterfs 2.0G 67M 2.0G 4% /media/gfs
To automatically mount the GlusterFS storage volume, you can update the
/etc/fstab with the line;
gfs01:/gfsvol01 /media/gfs/ glusterfs defaults,_netdev 0 0
_netdev ensures that the filesystem is not mounted until when the network is enabled and running.
In our next guide, we will discuss how to mount GlusterFS storage volume on a client using NFS.
That marks the end of our guide on how to install and setup GlusterFS storage cluster on CentOS 8.