Author Topic: Optimized Astropulse 5.03 for Windows (Read 26336 times)

Josef W. Segur · « **Reply #15 on:** 08 May 2009, 01:03:59 pm »

Quote from: Richard Haselgrove on 08 May 2009, 03:31:01 am

Gap-filling:

WU 440049768: ap_02ap09ac_B3_P1_00078_20090505_16821.wu

And upcoming: WU 441010635: ap_06ap09ac_B3_P1_00214_20090507_07720.wu
where task 1222567138 shows the problem remains.

Edit: Found in task list for host 4709496 (v16 Opteron):
WU 440332488 ap_07ap09ag_B3_P1_00121_20090506_31811.wu which has the problem, and
WU 436902732 ap_02mr09ac_B3_P1_00383_20090425_06576.wu which does not.
Joe

_heinz · « **Reply #16 on:** 10 May 2009, 01:19:01 pm »

Quote from: _heinz on 07 May 2009, 03:33:03 pm

Hi Joe,
As I'm going through the WU's I found 13 of them (B3P1) I will post when they are done.
they are from
02ap09ab ap_02ap09ab_B3_P1_00170_20090504_15112.wu http://setiathome.berkeley.edu/workunit.php?wuid=439903914
03ap09aa ap_03ap09aa_B3_P1_00250_20090505_10103.wu http://setiathome.berkeley.edu/workunit.php?wuid=440203658

this wu show no error
07fe09aa ap_07fe09aa_B3_P1_00258_20090405_02593.wu http://setiathome.berkeley.edu/workunit.php?wuid=430115086

13mr09ab ap_13mr09ab_B3_P1_00100_20090427_18772.wu http://setiathome.berkeley.edu/workunit.php?wuid=437453704
14mr09aa ap_14mr09aa_B3_P1_00143_20090428_28964.wu http://setiathome.berkeley.edu/workunit.php?wuid=437677134
14mr09ae ap_14mr09ae_B3_P1_00083_20090502_29770.wu http://setiathome.berkeley.edu/workunit.php?wuid=438809808
14mr09ae ap_14mr09ae_B3_P1_00030_20090502_29770.wu http://setiathome.berkeley.edu/workunit.php?wuid=438806928
14mr09ae ap_14mr09ae_B3_P1_00186_20090502_29770.wu http://setiathome.berkeley.edu/workunit.php?wuid=438813226
14mr09ae ap_14mr09ae_B3_P1_00000_20090502_29770.wu http://setiathome.berkeley.edu/workunit.php?wuid=438802784
15mr09ac ap_15mr09ac_B3_P1_00082_20090503_04267.wu http://setiathome.berkeley.edu/workunit.php?wuid=439201056
15mr09ac ap_15mr09ac_B3_P1_00095_20090503_04267.wu http://setiathome.berkeley.edu/workunit.php?wuid=439201094
15mr09ad ap_15mr09ad_B3_P1_00049_20090503_10075.wu http://setiathome.berkeley.edu/workunit.php?wuid=439363429
20mr09ab ap_20mr09ab_B3_P1_00366_20090430_19532.wu http://setiathome.berkeley.edu/workunit.php?wuid=438335691
20mr09ac ap_20mr09ac_B3_P1_00229_20090501_12024.wu http://setiathome.berkeley.edu/workunit.php?wuid=438441412
20mr09ae ap_20mr09ae_B3_P1_00046_20090501_18046.wu http://setiathome.berkeley.edu/workunit.php?wuid=438559553

Hi Joe,
the most interesting 03ap09aa is out now.
07fe09aa shows no error, very interesting.

heinz

Josef W. Segur · « **Reply #17 on:** 10 May 2009, 03:35:57 pm »

Quote from: _heinz on 10 May 2009, 01:19:01 pm

Hi Joe,
the most interesting 03ap09aa is out now.
07fe09aa shows no error, very interesting.

heinz

Some I've checked from other hosts seem to indicate all of February is OK, and so are March 1st and 2nd. There aren't any 'tapes' listed for March 3 through 8, March 9 and 10 still haven't been split yet, plus the aa one for March 12.

There should have been several hard disks full of data recorded between March 2 and 9, the ALFA receiver was scheduled for much observation. Perhaps the lack of 'tapes' indicates that's when they started having trouble writing the data to the removable hard drives. If so, maybe while trying to get that working they disturbed the cable from the detectors to the data acquisition card in the computer. Then the problem would show in the March 9 data. That's total speculation at this point, just a possibility from what Matt has said about the recorder, and what we're seeing.
Joe

_heinz · « **Reply #18 on:** 12 May 2009, 11:23:23 am »

Josef W. Segur · « **Reply #19 on:** 12 May 2009, 08:28:34 pm »

Quote from: _heinz on 12 May 2009, 11:23:23 am

I got a next serie of B3P1 with errors:
ap_06ap09ac
ap_22mr09aa
ap_22mr09ac
ap_22mr09ad
ap_22mr09ad
ap_22mr09ad
ap_22mr09ad
ap_22mr09ad
ap_22mr09ad
ap_22mr09ad
ap_22mr09ad

@ Joe, I will post this errors as long as it is necessary

heinz

I'm convinced all 12mr09 through 07ap09 work from B3_P1 will have the error. If you'll keep checking in case I'm wrong that would be good, specific reports aren't really needed otherwise. And when they split the 09mr09 and 10mr09 data, or something after 07ap09 appears, I'll be very interested in whatever we can learn about when the problem started and when/whether it ended.

I also see that 12mr09aa through 12mr09ae have been added to the set of available 'tapes'. The af and ag parts both had the problem, it would be fascinating if some of those earlier sections didn't. But catching the beginning of the problem that accurately is highly unlikely.
Joe

_heinz · « **Reply #20 on:** 14 May 2009, 02:10:04 pm »

Josef W. Segur · « **Reply #21 on:** 17 May 2009, 12:51:15 am »

Just to continue the record, the 09mr09 and 10mr09 files also had the problem on B3_P1. So the problem started sometime between the end of the 02mr09ae recording (which didn't exhibit the problem) and the start of 09mr09aa. That 02mr09ae could have been data from March 3, but hardly later since ALFA was being used fairly heavily.
Joe

_heinz · « **Reply #22 on:** 17 May 2009, 08:11:25 pm »

Richard Haselgrove · « **Reply #23 on:** 19 May 2009, 05:42:27 am »

I have an allocation for

ap_06mr09ad_B3_P1_00101_20090518_30837.wu http://setiathome.berkeley.edu/workunit.php?wuid=444843376

I've applied the compressibility test, and it is not compressible: the acid test will come when I run it in a day or two, but provisionally the start date for the stuck bit is pulled forward from 02mar to 06mar.

Josef W. Segur · « **Reply #24 on:** 19 May 2009, 10:55:39 am »

Quote from: Richard Haselgrove on 19 May 2009, 05:42:27 am

I have an allocation for

ap_06mr09ad_B3_P1_00101_20090518_30837.wu http://setiathome.berkeley.edu/workunit.php?wuid=444843376

I've applied the compressibility test, and it is not compressible: the acid test will come when I run it in a day or two, but provisionally the start date for the stuck bit is pulled forward from 02mar to 06mar.

Thanks! I'd gathered a short list of five 06mr09ad_B3_P1 WUs to monitor, was thinking of looking for some 04mr09ac samples too. The compessibility test is sufficiently diagnostic, I think.

If you get a chance, would you get the value out of the
<data_desc>
<start>
<time>
field and convert it to UTC? The OnlineConversion.com Julian Date Converter or similar will tell us the actual start time of the WU, being an 'ad' it could actually be March 7 or even 8 data.
Joe

lou snutt · « **Reply #25 on:** 19 May 2009, 11:46:13 am »

I've just returned a couple from 09mr09aa
and 10mr09aa

F.

Richard Haselgrove · « **Reply #26 on:** 19 May 2009, 12:36:09 pm »

Quote from: Josef W. Segur on 19 May 2009, 10:55:39 am

Thanks! I'd gathered a short list of five 06mr09ad_B3_P1 WUs to monitor, was thinking of looking for some 04mr09ac samples too. The compessibility test is sufficiently diagnostic, I think.

If you get a chance, would you get the value out of the
<data_desc>
<start>
<time>
field and convert it to UTC? The OnlineConversion.com Julian Date Converter or similar will tell us the actual start time of the WU, being an 'ad' it could actually be March 7 or even 8 data.
Joe

The raw time in that field is 2454897.5549821, which converts to Sat 7 March 2009 01:19:10 using onlineconversion.

Edit - I've given it a bump start, and it's running normally - as would be expected from the compression test.

Pappa · « **Reply #27 on:** 19 May 2009, 06:58:11 pm »

I have one that failed today and one that passed... A soon as the Forums are back I will post which they are...

BTW, I tried to consolidate the discussion to a single thread...
Edit
Okay
ap_12ja09ae_B3_P1_00110_20090307_15552.wu http://setiathome.berkeley.edu/workunit.php?wuid=422182744

Error
ap_21mr09ad_B3_P1_00134_20090510_28001.wu http://setiathome.berkeley.edu/workunit.php?wuid=441959393
ap_10mr09ab_B3_P1_00338_20090516_24490.wu http://setiathome.berkeley.edu/workunit.php?wuid=443962960
ap_12mr09ab_B3_P1_00026_20090516_23480.wu http://setiathome.berkeley.edu/workunit.php?wuid=444107319

Richard Haselgrove · « **Reply #28 on:** 20 May 2009, 09:18:21 am »

Quote from: Richard Haselgrove on 19 May 2009, 12:36:09 pm

Edit - I've given it a bump start...

Oooof - that was a crunch-and-a-half - "percent blanked: 82.71". Maybe the bit got stuck during the recording

?

I've preserved the datafile for http://setiathome.berkeley.edu/result.php?resultid=1231307250

Josef W. Segur · « **Reply #29 on:** 20 May 2009, 12:08:10 pm »

Quote from: Richard Haselgrove on 20 May 2009, 09:18:21 am

Oooof - that was a crunch-and-a-half - "percent blanked: 82.71". Maybe the bit got stuck during the recording ?

The other one I've seen completed, 1231344675, had "percent blanked: 87.79". That was about 15 minutes later [(170 - 101)*13.42 seconds] so doesn't refute the possibility that the beginnings of the problem are being seen. WU 444908104 is ap_06mr09ad_B3_P1_00399_20090518_30837.wu so right at the end of that channel. Could turn out to show the full problem...

An AP WU has 8 MiB of data and each byte can have any of 256 values. If it's normal data, the incidence of each possible value should be not too far from 32768. If the stuck bit problem has started, 16 of the patterns would have a much higher incidence than the other 240. I'm not sure what tool would be useful for that kind of check.
Joe

Author Topic: Optimized Astropulse 5.03 for Windows (Read 26336 times)

Josef W. Segur

Re: Optimized Astropulse 5.03 for Windows

_heinz

Re: Optimized Astropulse 5.03 for Windows

Josef W. Segur

Re: Optimized Astropulse 5.03 for Windows

_heinz

Re: Optimized Astropulse 5.03 for Windows

Josef W. Segur

Re: Optimized Astropulse 5.03 for Windows

_heinz

Re: Optimized Astropulse 5.03 for Windows

Josef W. Segur

Re: Optimized Astropulse 5.03 for Windows

_heinz

Re: Optimized Astropulse 5.03 for Windows

Richard Haselgrove

Re: Optimized Astropulse 5.03 for Windows

Josef W. Segur

Re: Optimized Astropulse 5.03 for Windows

lou snutt

Re: Optimized Astropulse 5.03 for Windows

Richard Haselgrove

Re: Optimized Astropulse 5.03 for Windows

Pappa

Re: Optimized Astropulse 5.03 for Windows

Richard Haselgrove

Re: Optimized Astropulse 5.03 for Windows

Josef W. Segur

Re: Optimized Astropulse 5.03 for Windows