Re: PATCH: pgbench - aggregation of info written into log

On Thu, Aug 30, 2012 at 3:25 PM, Tomas Vondra <tv(at)fuzzy(dot)cz> wrote:
> On 30 Srpen 2012, 18:02, Robert Haas wrote:
>> On Fri, Aug 24, 2012 at 5:25 PM, Tomas Vondra <tv(at)fuzzy(dot)cz> wrote:
>>> This patch is a bit less polished (and more complex) than the other
>>> pgbench patch I've sent a while back, and I'm not sure how to handle the
>>> Windows branch. That needs to be fixed during the commit fest.
>>
>> What's the problem with the Windows branch?
>
> Well, there are comments about how timestamp does not work on Windows
> (filling 0), and I'm not sure how that affects the patch (e.g. determining
> the aggregation interval). I have no Windows workstation available so I
> can't actually try that.

Hmm. That seems like it might be something we need to fix first...

>> Could you un-cuddle your brances to follow the PostgreSQL style, omit
>> braces around single-statement blocks, and remove the spurious
>> whitespace changes?
>
> OK, will do.
>
>> Instead of having both use_log_agg and naggseconds, I think you can
>> get by with just having a single variable that is zero if aggregation
>> is not in use and is otherwise the aggregation period. On a related
>> note, you can't specify -A without an associated value, so there is no
>> point in documenting a "default". As with your other patch, I suggest
>> using a long option name like --latency-aggregate-interval and then
>> making the name of the variable in the code match the option name,
>> with s/-/_/g, for clarity.
>
> Why --latency-aggregate-interval? It has nothing to do with latencies,
> it's merely number of transactions per interval.

Oh, I thought it was modifying the behavior of -l.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company