On Mon, 1 Aug 2016, Axb wrote:
On 31.07.2016 23:00, John Hardin wrote:
Folks:
It looks like we didn't get another successful weekly masscheck again,
even though if you check the counts today they are above the thresholds.
I suspect this is happening due to some results being submitted "late".
Imo, what would solve the problem is having more cpu cycles available
for the larger masscheckers and more independent masscheckers so the
delay doesn't depend on one or two fat processors.
Agreed, we need more masscheckers. We do have people who want to
contribute but their applications haven't been processed.
Sidney, have you had any success with the backlog of masscheck
applications from KAM? I suggested at your initial announcement of taking
over the reins from KAM on the dev list that you might want to make an
annoucement on the *users* list to resubmit masschecker applications
directly to you, but I didn't see that happen.
Mine is one of the fatter spam corpus which I will reduce to make it on time.
Will also try to move to a box with more cores or split to be able to
increase the amount of simultaneous jobs.
It might be very helpful to submit each result file as it is completed,
rather than submitting them all together at the end of processing. Looking
at the results details for the latest net pass, if that was done it looks
like your big ham corpus would be submitted at ~1117 rather than at ~2105.
That might be the smallest change we could make to the masscheck system
for the largest immediate benefit.
--
John Hardin KA7OHZ http://www.impsec.org/~jhardin/
[email protected] FALaholic #11174 pgpk -a [email protected]
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
Programming is an art form that fights back. -- Kevin S. Gallagher
-----------------------------------------------------------------------
4 days until the 281st anniversary of John Peter Zenger's acquittal
