On Jan 25, 2008 11:36 AM, David Brain <[EMAIL PROTECTED]> wrote:
> I'd appreciate some assistance in working through what would be the
> optimal configuration for the following situation.
>
> We currently have one large DB (~1.2TB on disk), that essentially
> consists of 1 table with somewhere in t
Hi David,
I have been running few tests with 8.2.4 and here is what I have seen:
If fysnc=off is not an option (and it should not be an option :-) )
then commit_delay=10 setting seems to help a lot in my OLTP runs.
Granted it will delay your transactions a bit, but the gain is big
considering
Matthew wrote:
On Fri, 25 Jan 2008, Greg Smith wrote:
If you're seeing <100TPS you should consider if it's because you're
limited by how fast WAL commits can make it to disk. If you really
want good insert performance, there is no substitute for getting a
disk controller with a good battery-b
On Fri, 25 Jan 2008, Greg Smith wrote:
If you're seeing <100TPS you should consider if it's because you're limited
by how fast WAL commits can make it to disk. If you really want good insert
performance, there is no substitute for getting a disk controller with a good
battery-backed cache to w
On Fri, 25 Jan 2008, David Brain wrote:
We currently have one large DB (~1.2TB on disk), that essentially consists of
1 table with somewhere in the order of 500 million rows , this database has
daily inserts as well as being used for some semi-data mining type
operations, so there are a fairly
On Fri, 25 Jan 2008, David Brain wrote:
The hardware storing this DB (a software RAID6) array seems to be very
IO bound for writes and this is restricting our insert performance to
~50TPS.
If you're seeing <100TPS you should consider if it's because you're
limited by how fast WAL commits can
Hi,
I'd appreciate some assistance in working through what would be the
optimal configuration for the following situation.
We currently have one large DB (~1.2TB on disk), that essentially
consists of 1 table with somewhere in the order of 500 million rows ,
this database has daily insert
