 |
| [TA]Assimilator1 2007-06-08 17:48:28 | When I got back from work today I found that both DPAD clients (C2D CPU) had stalled  ,the title shows the error messeges the clients were showing ,after the 'was' were a string of long numbers.What's this about? Btw I have screenshots saved of the various numbers. |
| Xanathorn 2007-06-08 17:59:36 | Usually it's due to a faulty result or a linebreak (empty line between results) in the results.dat or the samplefile. Try throwing them away and restart the client. |
| Xanathorn 2007-06-08 18:02:14 | Could also be an empty line in the results.txt I believe (in the case the client refuses to send the results to an ftp server and generates this error message)  . |
| [TA]Assimilator1 2007-06-08 18:06:39 | I pressed Y(es) multiple times & it then carried on ,but why would the results files get an empty line(s)? |
| Stephen Brooks 2007-06-08 21:26:06 | I launched a new lattice today, that could have had something to do with it. Ah, it could quite possibly have been that PhaseRotD got removed and replaced by PhaseRotDD so when you tried to send PhaseRotD results it no longer had a lattice to check up against. |
| [TA]Assimilator1 2007-06-09 13:00:56 | Oh ok ,so its nothing FUBARed on my new rig then What is this phaseRot stuff anyway?lol ,got a link about that? |
| Stephen Brooks 2007-06-09 16:42:06 | PhaseRot = Phase Rotation = Longitudinal Phase Space Rotation It's about the fact that the section "rotates" the bunch in energy-position space so that the particles end up with all the same energy instead of the front ones having the highest. I don't think I have anything uploaded about that area specifically. In fact, the PhaseRot designs simulate the target, decay channel _and_ the phase rotation. |
| [OCAU]DGROMS.com 2007-06-09 23:09:30 | Definately think this is the case Stephen as all 10 of my Quad core machines were stuck with checksum errors and the PhaseRotD file was missing on all machines. Could you not have left the PhaseRotD lattice file inplace and just added the extra file. I have lost hundreds of results due to this, I seem to remember a few years back this happening as well. |
| laurenu2 2007-06-10 07:30:37 | I to had 10 or so nodes that were stuck with checksum errors to over the past 10 ro 14 days |
| Stephen Brooks 2007-06-10 09:51:06 | DGROMS: I also remember it happening, but thought I got the client to do something different (e.g. download the new lattices, but don't delete any old ones, then send the results, then delete the old ones). Guess it needs a bit more work on it still. Overlapping the lattices on the server won't help because while D is up (regardless of whether DD is there), all your clients will be producing D results, which will then cause errors when it's removed. So maybe I could get the program to store all "old" lattices somewhere for checksumming purposes. |
| K`Tetch 2007-06-11 09:38:35 | you mean you thought you fixed in in 4.42c? its in the changelog as that being fixed, and i remember testing it out (i used to have every lattice file in the project) |
