I tried to connect the data pipeline not the read/write instance of my cluster but to the secondary, read-only one. The reason is quite simple: I want to offload the read/write instance - in my case a Mongo Atlas DB. But I found nothing neither in the documentation nor with CoPilot so I reached out to the support. And they told: It is useful, but not supported as of now. Could be an idea to make many customers happy 🙂
No CommentsBe the first to comment
Recent ideas
Adjust column width in SQL Analytics Endpoint user interface
I can't see the column names and cell values in some of the columns in my tables in SQL Analytics Endpoint because the column width is too narrow. Also, the tooltip also doesn't show the entire...frithjof_v1 day agoCommunity ChampionCompleted4.2KViews75likes8Comments