Our AWS EC2 Project
Migrating Bitrix24 Cloud to Bitrix24 On-Premise hosted on AWS EC2
After migrating Bitrix24 cloud with 1TB storage usage to on-premise and having allocated 3TB of EBS storage as part of the migration process, we found ourself with USD400 bill per month cost to host Bitrix24. We were only using 1TB of storage after migration but had 3TB storage on EBS and AWS was charging us 100GB daily which account to 3TB monthly.
After some research, we decided that we need to shrink our EBS volume and use AWS S3 storage which is way cheaper. So, here are the steps we took to shrink our 3TB EBS volume to 850GB.
You may want to try the steps using AWS free tier account first using a Linux EC2 instance of 20GB and shrink it down to 10GB volume.
Take a snapshot of your existing EBS volume
This is for backup and also for you to create a new volume in which you will shrink. A 3TB volume takes about 2 hours to complete and be available.
If your Bitrix24 instance is already in production mode, do inform your users that anything they perform in system onwards will not be available after the scheduled maintenance.
Create a new volume from snapshot
GP3 volume type is recommended as it's much cheaper. Make sure you create in the correct sub-net and region.
Launch new lightweight AMI instance
You can use Bitrix24 AMI -BitrixVM-7.5.1_CentOS7.9_x64_HVM_EBS
Attached the new volume to this new instance. The volume should be attached as xvdf.
Resize Filesystem
SSH to your terminal with no timeout. Use tmux to run the command so that you can remote terminal using your mobile phone to check progress. You can use the free Termius IOS app. Run below command.
sudo e2fsck -f /dev/xvdf3
sudo resize2fs -M -p /dev/xvdf3
A resize of 3TB volume takes about 12-14 hours.
Create new smaller EBS Volume (GP3)
Copy data from shrink volume to new small volume
Run sudo dd bs=16M if=/dev/xvdf3 of=/dev/xvdg3 count=50000 status=progress
[Change 50000 to your own counts]
Initialize instance with new smaller EBS Volume
sudo resize2fs -p /dev/xvdg3
sudo e2fsck -f /dev/xvdg3oo
Detached from temporary instance.
Shutdown production instance and detached volume
Attached new smaller volume to production instance and start the server.
