Hi,
Due to a planned migration of date from an old system, our data is going to grow.
Currently we are not using shards, and all our data is using one server (actually 3 replicas, but no sharding).
I was wondering if there is a rule of thumb on the size limit of data on a single shard/server.
In other words - what is the max size of my data that requires using more than one server and using shards?
Thanks,
Tamar
Hi @Tamar_Nirenberg,
I remember up to 2 TB you can mantain a repl set, but from the production notes, is a bad way to mantain a replica set only in one node!!!
in attachment, the link with the details from the documentation:
https://www.mongodb.com/docs/manual/core/replica-set-architecture-geographically-distributed/#:~:text=While%20replica%20sets,centers%20is%20unavailable.
If I retrive the post with this information, i Will post It!
Regards
By
and
Do I understand correctly that you are running 3 instances of the replica set on the same machine? Or do you have 3 machines.
Except for experimentation, there is not point in running 3 instances on the same hardware. This is a single point of failure. If you can live with a single point of failure you are better off with 1 mongod instance.
Sorry, I guess I did not make my self clear.
I have 1 Primary, and two standby’s, each running on a different server.
They all have, of course, the same amount of data.
My question is how large can the data be before I need to consider sharding the data.
Thanks,
Tamar
Sharding is powerful but also challenging to maintain. , generally you can always try vertical scaling first until you get to the ceiling before you start horizontal scaling (i believe this is also recommended by many books/sites). This is more a thing to ask your team than to mongo team.