We’ve observed that RAM usage on some of our Couchbase server nodes reaches approximately 80% of the total quota. Can we configure the server to exclude certain document types from loading into RAM, opting instead to retrieve them directly from disk upon request?
Thank you.
That’s normal. Good, even.
You could put them in a different bucket that has very little RAM.
That’s normal. Good, even.
You could put them in a different bucket that has very little RAM.
we keep getting a warning about it when the web console is accessed anywhere and moreover it doesn’t align with the access patterns of our data. The data is (and needs to be) part of the same collection, but relatively higher in size. So, we were thinking that controlling its loads would be a good way to control the RAM consumption. Do you mean there are no configurations/controls to do these things in couchbase?
Can you copy/paste the message? Thanks.
Is there a feature that accepts a list of document ids, and for those documents, removes them from RAM immediately after performing an operation? None that I know of.
What makes it necessary for these documents to be in the same bucket?
Thanks for your quick responses @mreiche. Much appreciated.
We use couchbase mobile and the data needs to be synced to the mobile clients via the sync gateway. Separating the content which is part of the same app into multiple buckets and collections will significantly increase the complexity in the sync gateway layer (assuming it is possible to sync multiple bucket data onto mobile clients).
And what’s the exact text of the message?
I’ll paste the message here next time I see it.