Sorry for my newbie questions, it is because I'm trying to find out how a Enterprise Storage like NetApp can aggregate values to mongodb implementations. It is very important to me understand how analyze I/O at MongoDB. I'm wanna be a MongoDB Database Administrator. :-)
Is there any way to see which collections are mostly accessed?
Hi Rodrigo,
The previous links I provided should be a good starting point for monitoring & stats documentation.
A few items which might be of particular interest for you from those pages are:
MongoDB uses memory-mapped files for disk I/O, so the general performance metrics on I/O, Ram, etc can be monitored using your normal O/S admin tools.
You can query the profile data to get some detail on access patterns by db+collection ("namespace"), but the information is probably more detailed than what you are after.
There are a few community-contributed tools that can work with the profiler data and help make recommendations on indexing:
Profiling uses a capped collection so will keep a limited amount of history; you can increase the size if needed:
You mentioned you are based in Brazil. FYI, there is also an upcoming user group meeting in Sãu Paulo at the end of the month which will include a few visiting 10gen team members:
http://www.meetup.com/SP-MongoDB/
I would suggest you set up a local instance of MongoDB to test with (if you haven't already!).
Cheers,
Stephen