Max file limits
I'm planning to implement a CouchDb database that is likely to have thousands of very small databases. I'm planning on using the standard railway install with volume attachment, but I'm wondering if I need to engineer a way to cluster them out or not (yet) based on the limitations of this structure. Since each database has a separate file (at least that's my understanding), this can cause issues in unix systems with max files in a directory/max files accessed at once. I'm not expecting a ton of databases to all be open at one time, but it might be close to 500ish.
I don't think it should cause issues, but I wanted to ask from an infrastructure standpoint if it's likely to hit resource limits.
This is the documentation page on resource limits from CouchDb: https://docs.couchdb.org/en/stable/maintenance/performance.html
Not sure how this works internally so if someone who's more familiar with this server setup can help me out, that'd be awesome!
5 Replies
Project ID:
438c7dbc-75c8-4f44-9268-d08b016ec3d8438c7dbc-75c8-4f44-9268-d08b016ec3d8
As I thought about this, I decided that I needed to implement this man-in-the-middle anyway, so you can mark this question solved. I would still like to know a little bit about the file access limits if possible so I know how often I should create a new item in the cluster.
railway does not impose any file limits in this regard
Okay, thank you. BTW, I see you were added to the team, congrats Brody!
thank you!!