Optimised Operational Concept for the EGroupware Hosting
Dear EGroupware users,
Quite a bit has happened during the last weeks. As already announced, the whole hosting cloud infrastructure moved to the German data centre provider Profitbricks.
New Data centre and optimised Operational Concept
The move is successfully completed, all instances are now running in the new data center in Frankfurt/Main, Germany. After an extensive research and talks to different providers, we decided to work with Profitbricks as an experienced, German cloud computing supplier. We’re happy to present you the new operational concept for more reliability today.
Standards
- Data storage takes place in Frankfurt in two independent storages in different fire compartments on several hard drives.
- Servers are virtualised and will be automatically started on new hardware in case of a hardware failure.
- Servers can be equipped with more ressources anytime during operation (e.g. storage, CPUs).
- The EGroupware hosting farm runs under the open-source system “Kubernetes”. It manages the orchestration of the containers. The hosting farm itself communicates with a “Galera” database cluster and the “NFS” server for files.
- Daily database backups and snapshots of the files are automatically created and saved for at least 30 days. This means, if you accidentially delete files, they can be restored.
The standards of Profitbricks and EGroupware guarantee an extended stability and failure safety for all EGroupware hosting customers.
Technical Setup of the Hosting:
Permanent Replication of all data in a second data centre in a different location
To offer even more failure safety, EGroupware now additionally runs in a second Profitbricks data centre in Karlsruhe, Germany. All EGroupware data is consistently replicated (files every 15 minutes) and stored in this second data centre. In case of a service failure in Frankfurt, hosting customers can hence access their latest data in Karlsruhe. In this case, we’ll automatically activate a reading access to your data. Should there be a longer service failure, we now also have the option to enable the backup hosting farm as a primary platform. Even in case of a longer regional interruption, we can therefore react quickly and make all your data available.
Our Graphic explains the functionality of the new operational concept:
Testing of the Backup Hosting Farm:
To test your EGroupware data, stored in the backup farm, you can access it with this URL anytime: „instance name – top-level domain.egroupware.cc“.
For example for our support centre instead of “my.egroupware.org”: „my-org.egroupware.cc“ or “mycompany-net.egroupware.cc” for all “.net” domains. In case of a service failure, the switch from the primary hosting farm to the backup hosting farm takes place automatically.
Preview – Move of the Mail Server:
The scheduled move of the EGroupware mail farm to the data centre of Profitbricks will take place by the end of June. We’ll also establish a replication of all mail data in both data centres. Therefore we’ll also achieve an improvement of spam filtering and virus protection. Customers of the EGroupware mail hosting will be informed about this process in a separate e-mail.
Kind regards
Birgit Becker, EGroupware GmbH
