Final Question 3 - SynchingTo:

Hello

While looking at the example from the rs.status().members() I noticed that the syncing arrangement was not consistence.

27017 Primary
27018 syncingTo: 27022
27019 syncingTo: 27022
27020 syncingTo: 27022
27021
27022 syncingTo: 27017
27023 syncingTo: 27022

I was wondering why this unusual syncing chain - Primary <= 27022 <= 27018, 27019, 27020, 27023

  1. So is it anything I need to concern myself with when answering the question? :slight_smile:
  2. I appreciate syncingTo has been depreciated from since 3.4 +, I have to question why it was called syncingTo instead of syncingFrom ?

Hi NMullins,
I wondered that myself, then took a look at my own repl set. I had
20004 (secondary) syncing to 20006 (secondary) syncing to 20005 (primary).

if you think about it, this is a good thing. it means that a secondary can sync with another secondary. It avoids contention on the primary with every node trying to sync with a single node.

I choose to ignore the syncingTo: for this reason.

Garrett

1 Like

A good point.

I had chosen to ignore it, but now with a more confidence. I did jump onto a replica set to have a look and it was as you said.

As they say, assumption is the mother of all โ€ฆ