Batching Datas From Hidden Secondary Node

Hi All,

We currently use cloud manager for maintain our mongodb . so we have plan for build data extraction from spesific Hidden Secondary Node in production . is it safe ? is it ok if we run it in every hour ?

Hi @Abdul_Haris welcome to the community!

“Safe” is a relative term here, and I don’t think Cloud Manager comes into effect in this case (it is a management tool after all).

Typically a hidden secondary is used to perform tasks not associated with the replica set’s usual workload, so client’s won’t see it, and won’t route queries to it even when secondary or secondaryPreferred read preference is used.

However it still behaves like a secondary, so all the caveats regarding a normal secondary still applies:

  • It still does as much work as a primary and other secondaries to keep up with the state of the primary
  • It still relies on not falling off the oplog to be part of the replica set

Thus, the suitability of the hidden secondary to do what you asked for is dependent on how much work you’re asking it to do, on top of its regular work of keeping up with the primary. If it can handle it, it’s should be ok. If it cannot, then you might see some issues with it.

I would suggest you to experiment with some example workloads to start with, and see if the hardware can keep up with demand.

Best regards