I didn't understand a problem. Can the files that have been lost be recovered over a certain period or are they permanently lost and should we start from 0?
When we’re done with all necessary tasks, we will recover most important servers. Updating the security settings of an old server takes more than 15 hours due to the security files size, while adding new server takes 15 minutes, that’s why we’re adding new ones instead of updating the old ones for now.
15 Hours in 200 server will take 3,000 hours, while adding new 200 servers will take 3,000 minutes.
However, in order to avoid this in the future we will build a new system with 3 levels of external backups that works in this order:
-
OVH as
Front-End (Streaming)
-
Online SAS as
Online Backup (First level, that works in case of OVH failure)
-
FDC as
Online Backup (Second level, in case Online resources wasn’t able to handle all the load)
- Leaseweb as
Offline Backup (this backup will never be online, but only to import our partners files from it in case we lose the front end, first and second levels. The offline data will never be exhibited to any risk considering its offline.
Previously, we used to have only OVH as front end and Zain Telecommunications as backup, which means one level instead of 3. The Zain Telecommunications servers speed when we tested them is less than 100 Mbps per each, which means we need a lot of time to restore all files and that’s why I asked most users to re-upload instead so they can reduce the number of files we have in Zain Telecommunications with their lovely speed.
The new system we’re building will never fail unless WW3 starts. The secret of the new system is not the 3 new levels, but the
offline backup.
We may be able to do 10 level of backups, but that would be risky if all of them are online because if one of them got hacked the rest will be as well, but the offline backup will never receive any external commands, but only users files through specific local port.
