Database Server Speed

DB as oracle…some of my preliminary observations -

  1. As started working on reports on upon db server, my development team facing speed issue as they are also working upon another part of development within same database.
  2. As I load dashboard when I login the metabase, it start processing reports (may be compiling the reports). If I built question using native query, usually it may involve more than 2-3 tables. These questions do appear on my dashboard as my client may need them.
    My questions are
    a. If the questions are compiled on every start or refresh(auto) then do the tables concerned gets locked for that time (my data contains more than 100k rows/records approx for transaction tables) ?
    b. Do the performance in future, I mean transactions performed by the users of the system will experience this problem if i created multiple dashboards and they are loaded by their supervisors within office hours(fixed time frame) …?
    And will it be faced by supervisors of the system who use their dashboard for analysis purpose?


Hi @mganorkar

If your database is overloading, when you are running Metabase, then you should consider running analytics on a slave instance that is read-only - this is also recommended when doing backup - so tables on the master instance are not being locked.

Metabase has a cache function, which you can enable, so visits to the same query are saved for X seconds/minutes. That should help, if a lot of people are viewing the same dashboards/questions.

Metabase is only as fast as your database. If your database is slow, then Metabase will feel slow.

1 Like

Hi Flamber,

Thanks a lot for your prompt reply!

– Can you please put light upon "you should consider running analytics

on a slave instance"? Where I will get this option?

– What will happen when MAX CACHE ENTRY SIZE exceeds more than the

Kilobytes mentioned in MAX CACHE ENTRY SIZE ?

(is there any way to like/subscribe/suggest/recommend the replied?)



Master/slave replication means you synchronize data over multiple database server, and then you use the slave instances for running heavy read-queries like analytics or backups.

When you exceed the max cache entry size, then Metabase will not cache the result, so next time the question is being run, it will execute the query on the database instead of trying to find an existing cache.
If you set this setting too low, then it’s almost like disabling cache.

1 Like


Thanks a lot! Yes, this seems inspiring! :+1::+1::+1: