administrators should somehow be able to check the current (and history) system
performance. sample data that should be stored might be:
- number of method calls (especially for methods provided by the rpc service)
- the time it takes to do n requests on the server (n=1 and n=100 for example)
- the time it takes to do n requests from the client to the server and
receiving the answer (distinguish between read- and write-intensive tasks;
display min, avg, max values)
- an overview about the current number of distinct values for each column of an
entity -> to identify potentially unused columns. goal: keeping the number of
columns as small as possible.
Original issue reported on code.google.com by ingo.jae...@gmail.com on 14 Jul 2010 at 3:23
Original issue reported on code.google.com by
ingo.jae...@gmail.com
on 14 Jul 2010 at 3:23