New Manual Results Submission Page
 

When submitting a larger amount of results, the manual results page feels a lot slower than it did before. For one thing, nothing gets displayed until the batch is completely processed or timed out.

I personally preferred the results slowly pouring onto the page so you could see progress (and relative speed). I remember being able to push through ~80 KB in one go but now I can't. I'm going to reduce the batch size until I find a size that works better.

Has anyone else felt that the page is slower?

As of 2:30 GMT-6:00 the manual results is going very poorly. After submitting a file much smaller than what I used to submit (now down to ~50 KB) the page works for about 3 minutes and has enough time to process about sixty lines, which is maybe 5KB worth.

I also find it rather slow. Don´t have hard figures but in general it´s going slower than it used to be, specially when uploading files (pasting results to the window seems to go at roughly the same speed). But it also varies with the moment chosen to check in the results. Sometimes the server is more heavily loaded, and that is apparent in the speed at which results lines are "swallowed" by the server.

Yeah. Nine hours later I was able to push through a 120 KB file in a swallow and a half. No idea why it was so bogged down at 2:30 in the morning. I looked at the recent cleared and recent results and I found absolutely nothing to indicate the server was being otherwise stressed.

It could have been under stress from something else, however.

I was having some slowdowns while manually submitting TF work over the past few days with even 4-5 Kb files taking 30 second or so sometimes more(note my internet speed not problem) But the submissions page seems back to normal as it only takes a few seconds to process files now.

I think this is a "late night" problem. I'm currently submitting my files in little 5KB chunks. Whenever I do this earlier in the day, it works great. Whenever it's later in the evening (even 10PM GMT-600) it's slow as hell.

Does anyone know why this is a thing? I can't understand why the server would see so much traffic at this time of night that it can't process a few KB of data.

Maybe someone who understands can devote a minute of their time to explaining to me why the server missing 95% of its processing capacity over what is clearly a serious amount of time isn't a big deal.

If you install Perl you can run [URL="http://www.gpu72.com/spider/"]this script[/URL] hourly and send work in slowly but steadily.

[url]http://www.mersenneforum.org/showthread.php?t=16334[/url]

It works very well and we have never lost any results with it.

[QUOTE=Xyzzy;369929]If you install Perl you can run [URL="http://www.gpu72.com/spider/"]this script[/URL] hourly and send work in slowly but steadily.

[url]http://www.mersenneforum.org/showthread.php?t=16334[/url]

It works very well and we have never lost any results with it.[/QUOTE]

Ah yes.. I almost forgot about that :razz:
It's useful mostly for linux but I don't think many windows users use it because of misfit.

I noticed the following problem using the "Results" in "Manual testing":

If I use a result file (a subfile of my result.txt), then all F-PM1 results had not been recognized (but all NF-PM! results)!

Only after deleting the
(P-1 B1=635000, B2=13176250, E=6)
resp.
(P-1 B1=730000)
phrase, the factors get accepted.

This is not very comfortable...

Regards,
Jimmy

[QUOTE=JimmyB;370155]I noticed the following problem using the "Results" in "Manual testing":

If I use a result file (a subfile of my result.txt), then all F-PM1 results had not been recognized (but all NF-PM! results)!

Only after deleting the
(P-1 B1=635000, B2=13176250, E=6)
resp.
(P-1 B1=730000)
phrase, the factors get accepted.

This is not very comfortable...
[/QUOTE]

Can you post the exact text that the web page failed to recognize? Thanks.

[QUOTE=Prime95;370630]Can you post the exact text that the web page failed to recognize? Thanks.[/QUOTE]

The text of the input file was:

[Sun Mar 30 22:04:30 2014]
UID: jimmyb/iMac3, M65987773 completed P-1, B1=635000, B2=13176250, E=6, We10: ABCC461E, AID: 183CBD9B58C077CF10C830D0E7DB0FB3
[Mon Mar 31 13:17:38 2014]
P-1 found a factor in stage #2, B1=635000, B2=13176250, E=6.
UID: jimmyb/iMac3, M65988487 has a factor: 2561952789475879468530426891847 (P-1 B1=635000, B2=13176250, E=6), AID: 88177E712525EDE2BFAD93FB1D55B5C4
[Mon Mar 31 23:16:02 2014]
P-1 found a factor in stage #2, B1=635000, B2=13176250, E=6.
UID: jimmyb/iMac3, M65987743 has a factor: 189351851647586625240525399298681 (P-1 B1=635000, B2=13176250, E=6), AID: E4BA33419EFF84F893A24E518B57BD94
[Wed Apr 2 02:57:25 2014]
UID: jimmyb/iMac3, M69851273 completed P-1, B1=730000, B2=16972500, E=6, We10: 0E1993F7, AID: 40524F75529202E2A501FF88911DF0B6


The result lines were produced by
prime95 v28.3

Regards
Jimmy

Should be fixed. The manual results page did not handle both the (P-1...) and AID strings.