[Sqlgrey-users] Testing a middle-size to large SQLgrey installation

[Jean-Michel Pouré - GOOZE]

Dear all,

I would like to build a middle/big size SQLgrey database for testing,
in a situation of a middle-size ISP.

My first settings show that for each user, 50 to 100 records are added
everyday to the database. Records are added and removed, so we need to
monitor insertion, deletion, query.

If I consider 1 million users, 1 day to reconnect, it is a database of:
1.000.000 x 50 = 50 million records added and deleted everyday.

Is this the kind of size that SQLgrey is facing? If this is the case,
we probably need very special indexing (or no indexing at all).

In your opinion, what is the size of a large SQLgrey database?

Kind regards,
Jean-Michel

[k***@rice.edu]

Post by Jean-Michel Pouré - GOOZE
Dear all,
I would like to build a middle/big size SQLgrey database for testing,
in a situation of a middle-size ISP.
My first settings show that for each user, 50 to 100 records are added
everyday to the database. Records are added and removed, so we need to
monitor insertion, deletion, query.
1.000.000 x 50 = 50 million records added and deleted everyday.
Is this the kind of size that SQLgrey is facing? If this is the case,
we probably need very special indexing (or no indexing at all).
In your opinion, what is the size of a large SQLgrey database?
Kind regards,
Jean-Michel

Hi Jean-Michel,

50m adds/deletes = 100m transactions/day which works out to 1157 per sec.
Even with a standard hard drive, with enough memory and with sync_commit
off, this is easily manageable for a completely in-memory DB and it is
not even large.

Regards,
Ken

------------------------------------------------------------------------------