Hi Dan! Only positive results so far. After half a year! Just yesterday I increased oplog size on my mongodb and restarted database in the midst of peak traffic, no downtime, no problems. It’s so awesome to have mongodb up and running, and spot instances fit just fine:
Driver, mongos and mongodb - together they make it unnoticeable any server termination as another node quickly becomes primary and the system is available for writes in no time.
Here my short LinkedIn post: Sergey Zelenov on LinkedIn: #mongodb #googlecloud #kubernetes #costefficiency
Would you mind to add a reaction, I’d love to share our experience with other mongodb users 