Standalone Backendless with DB cluster/replica set ??

Standalone Backendless
1

Hi -

We are working to see how much Backendless fits our requirements, in terms of production architecture - so, request for your help with below queries - in regard to standalone deployment:

  1. Does the backendless use MongoDB or MySQL for internal database ??

  2. Can we configure the internal DB of Backendless as a DB cluster / replica set ?? If yes, does Backendless instance manages/helps in the same in any way - or we need to think DB as separate layer, and follow process as per the particular DB type ??

  3. Can we access internal DB directly, for automated backup purposes ??

  4. Can we configure Backendless as a 2 node cluster (active-active) ?? If yes, can the internal DB of these 2 nodes, act as DB slave for each other (this avoiding separate VMs for DB slaves) ??

  5. Can we use MySQl/MongoDB instance created by Backendless - to store our own separate schema of tables ?

Thanks.

Regards,
Aditya

2

Hi Aditya,

Backendless uses MariaDB database, which is a fork of MySQL. In Standalone you have complete direct access to the database as it is installed in a separate folder withing your whole Standalone installation. This means you may configure it any way you like, do manual or automatic backups, store any additional data you want, and Backendless will work properly as long as the database has one stable address.

Currently Backendless server can’t work properly with master-slave replicated database (read from slave, write to master etc.), but we are actively working on it now.

Also remember that free installation is limited to only one machine instance.

3

Hi -

So this means, I cannot setup either replication or cluster of MariaDB - and get it working with Backendless.

What does Backendless then propose, to safeguard data against any HW or SW failure ??

My requirement is simple - once the data is committed to DB, there should be any single point of failure that may result in loosing the data.

4

No, I didn’t say that you cannot set up replication or cluster. The only requirement Backendless has for database is that it has one defined address (host:ip). If you configure your database to be replicated to other instances - it will work just fine.

5

ok …so we can have VM1 running Backendless(with master MariaDB) - and VM2 running slave MariaDB.

How about 2nd Backendless instance - Can 2 Backendless work in Cluster ?? Also, how does the DB synchronization will work in that case - as both Backendless will have their own master Maria DB ???

6

Multiple Backendless instances may work in cluster in case you have some kind of load balancer before them. But note that running multiple Backendless instances is not free - https://backendless.com/pricing/

Also you will need to have one entry point (host:port) in your database cluster (this means you would need to configure database cluster on your own).

7

yes, I know the pricing part - have already discussed it with Edward.

So, the 2 instances of backendless wont be talking to each other - and will work as cluster only due to external ‘load balancer’ and ‘common DB’ ??

Will something like in attached image work fine ??

backenless_setup.jpg
backenless_setup.jpg1280×720 77.8 KB

8

Yes, what you have in the diagram will work.

9

Aditya

Did you go with backendless

Did you get this configuration working

10

Hi Barry - I know it was an old post - but our production is still couple of month’s away - and we haven’t yet tried this config in lab yet. I am hoping it will work.