DB connection times
dochawk at gmail.com
Mon Mar 2 23:08:11 EST 2015
On Sun, Mar 1, 2015 at 9:27 AM, Richard Gaskin <ambassador at fourthworld.com>
> Dr. Hawkins and others have noted that the time required for LC to connect
> to a database can be long enough to add up under load.
> To get a good feel for the implications of this, and to explore options
> for possibly mitigating it, it would be helpful if any of you happen to
> have measurements of that connection time.
I think it might make sense to make a standardized test suite for us to
run, so that we're comparing apples to apples.
In my case, I know that latency is far more important than throughput, as I
need the transaction to occur during using input.
As a rough, from a fast link at home (10Mbs?), it takes "several" seconds
for the series of transactions it takes me to open a client table, compared
to "a second or two" on my desktop (which hosts the server) on a good
Dr. Richard E. Hawkins, Esq.
More information about the Use-livecode