/
Accelario Server High Availability

Accelario Server High Availability

Accelario server HA on AWS - using EBS

  1. On EC2 instance of Accelario server add EBS drive (ext4) for accelario server binaries and mount it under /accelario/loki

   In the /accelario/loki folder, put the server binaries.

  1. Add block devices for ZFS pool

  2. Use VIP for attaching static IP to the server

  3. Prepare a second standby server for Accelario server with all required packages

When Accelario server goes down you will need:

  1. Make sure the EC2 is down

  2. Start the second Accelario server

  3. Detach EBS devices from the failed server

  4. Attach the EBS devices to the new accelario server

   a. attach all zfs pool devices

   b. attach the EBS volume which contains Accelario server binaries

      b1. This device will have to be mounted under same directory as it was on the previous machine. Best practice is to add /etc/fstab with UUID for this device.

  1. Attach VIP to the new accelario server

  2. Run zpool import -a command on the new accelario server, it will import all zfs devices metadata.

  3. Start the accelario server on the new EC2 machine

Accelario server HA with multi AZ disaster recovery on AWS - using EBS

  1. On EC2 instance of Accelario server add EBS drive (ext4) for accelario server binaries and mount it under /accelario/loki

   In the /accelario/loki folder, put the server binaries.

Setup automatic snapshot for this EBS drive

  1. Add block devices for ZFS pool

  2. Use VIP for attaching static IP to the server

When Accelario server goes down you will need:

  1. Make sure the EC2 is down

  2. Start the second Accelario server from the EBS snapshot

  3. Attach all zfs pool devices

  4. Attach VIP to the new accelario server

  5. Run zpool import -a command on the new accelario server, it will import all zfs devices metadata.

  6. Start the accelario server on the new EC2 machine

Accelario server HA with multi AZ disaster recovery on AWS - using EBS and FSX

  1. On EC2 instance of Accelario server add EBS drive (ext4) for accelario server binaries and mount it under /accelario/loki

   In the /accelario/loki folder, put the server binaries.

Setup automatic snapshot for this EBS drive

  1. Add FSX volumes as a storage pool

When Accelario server goes down you will need:

  1. Make sure the EC2 is down

  2. Start the second Accelario server from the EBS snapshot

  3. Start the accelario server on the new EC2 machine

  4. Attach the same FSX volumes as a storage pool

In case of human errors - use the same solution:

  • For the binaries - restore from a snapshot of the correct time (before the human error)

  • For the data - use the snapshots that are built in the application

WhatsApp Image 2025-10-17 at 14.21.30_f4b16636-20251017-112130.jpg

 

Accelario server HA on ESXI

  1. On VM instance of accelario server add VMDK drive (ext4) for accelario server binaries and mount it under /accelario/loki

   In the /accelario/loki folder, put the server binaries.

  1. Add block devices for ZFS pool from ESXI storage

  2. Use VIP for attaching static IP to the server

  3. Prepare a second standby server for Accelario server with all required packages

When Accelario server goes down you will need:

  1. Make sure the VM instance is down

  2. Start the second Accelario server

  3. Detach VMDK devices from the failed server

  4. Attach the VMDK devices to the new accelario server

   a. attach all zfs pool devices

   b. attach the vmdk volume which contains Accelario server binaries

      b1. This device will have to be mounted under same directory as it was on the previous machine. Best practice is to add /etc/fstab with UUID for this device.

  1. Attach VIP to the new accelario server

  2. Run zpool import -a command on the new accelario server, it will import all zfs devices metadata.

  3. Start the accelario server on the new VM machine