v4_computer vs. Machine Name anomaly

[ADBjester]

Hello All,

I migrated when migration still had some bugs. Short version:

I let v24 run a few days and report to the v5 proxy.
I stopped v24, upgraded the software to v25.7, edited the worktodo.txt.
For whatever reason, it ignored the worktodo.txt and acquired new assignments.
Bugs in the server were fixed.
I stopped v25 again, re-edited worktodo.txt, and eventually my old assignments showed back up in "assignments":

v4_computers 0 39142601 LL 40.50% 2008-10-27 09:17 13 2008-10-31 03:18 2008-11-01 03:18 2008-11-18 03:18 9
v4_computers 0 39826379 LL 46.10% 2008-10-27 02:47 13 2008-10-29 18:27 2008-10-30 18:27 2008-11-15 18:27 6
v4_computers 0 42126521 LL 74.30% 2008-10-27 02:47 13 2008-10-29 18:27 2008-10-30 18:27 2008-11-06 18:27 -3
v4_computers 0 42241393 LL 52.40% 2008-10-26 23:47 14 2008-10-29 18:26 2008-10-30 18:26 2008-11-14 18:26 5
v4_computers 0 43060733 LL 93.90% 2008-10-27 23:06 13 2008-10-31 03:19 2008-11-01 03:19 2008-11-03 03:19 -6
v4_computers 0 43713101 LL 19.60% 2008-10-27 17:48 13 2008-10-29 18:27 2008-10-30 18:27 2008-12-02 18:27 23
v4_computers 0 44036719 LL 94.90% 2008-10-27 04:40 13 2008-10-31 03:19 2008-11-01 03:19 2008-11-03 03:19 -6
v4_computers 0 44106793 LL 86.30% 2008-10-27 10:32 13 2008-10-31 03:19 2008-11-01 03:19 2008-11-07 03:19 -2
PrestONE 1 46564261 LL S2, 8.30% 2008-10-29 18:46 11 2008-11-08 04:12 2008-11-09 04:12 2008-12-20 22:36 41
PrestONE 2 46564303 LL S1, 5.50% 2008-10-29 18:46 11 2008-11-08 04:12 2008-11-09 04:12 2008-12-29 12:05 50
PrestONE 3 46564327 LL S1, 6.00% 2008-10-29 18:46 11 2008-11-08 04:12 2008-11-09 04:12 2008-12-27 21:22 48
SkinnyPuppy 2 46666027 LL LL, 6.50% 2008-10-31 21:35 9 2008-11-08 03:29 2008-11-09 03:29 2008-12-31 23:45 52
SkinnyPuppy 3 46675087 LL LL, 5.40% 2008-11-01 03:27 8 2008-11-08 03:29 2008-11-09 03:29 2009-01-01 14:52 53
SkinnyPuppy 0 46874717 LL S2, 41.30% 2008-11-04 21:29 5 2008-11-08 03:29 2008-11-09 03:29 2009-01-04 23:03 56

All well and good.... but there are a couple of problems:

1) Because there was still quite a ways to go on the older exponents that were in progress, I manually stopped v25, REMOVED the new assignments from worktodo.ini (some of the ones shown above, as assigned to machine names on 10/29 and 10/31), and then tried to use the "Assignments" page to 'Unreserve checked exponents'. I can't seem to do that. No biggie -- I'll eventually get to those exponents when the older work started by v24 finishes up.

2) That brings me to my second issue. As work started by v24 machines finish up, the results are reported back to the server... but they are reported UNDER THE V5 MACHINE NAME rather than as a v24 v4_computer. Several of the exponents shown above today in my assignments list have actually been completed and reported as composite:

PrestONE 42126521 C 2008-11-07 08:44 0.0 C93BEF1C001C3B__ 73.1363
SkinnyPuppy 46675087 NF-PM1 2008-11-07 00:53 5.9 B1=565000, B2=15820000 2.8446
SkinnyPuppy 46666027 NF-PM1 2008-11-06 22:34 6.0 B1=565000, B2=15820000 2.8446
SkinnyPuppy 44106793 C 2008-11-06 10:01 0.0 1371E984DB5FB0__ 76.5743
SkinnyPuppy 43060733 C 2008-11-02 16:59 0.0 086EC486654461__ 74.7582
SkinnyPuppy 44036719 C 2008-11-02 08:03 0.0 B94444139A5DF0__ 76.4526
v4_computers 43713101 NF-PM1 2008-10-27 17:48 0.0 2.4036

No matter what I do, I can't seem to remove these exponents from my "assignments" list (as v4_computers), even though the result has already been turned in. Thus, as you can see, some of my "assignments" are not aging properly.

Is there any way to safely drop these completed assignments from my v4_computers assignment list, without losing credit for them?

Thanks!

Jester

[petrw1]

Originally Posted by ADBjester

1) Because there was still quite a ways to go on the older exponents that were in progress, I manually stopped v25, REMOVED the new assignments from worktodo.ini (some of the ones shown above, as assigned to machine names on 10/29 and 10/31), and then tried to use the "Assignments" page to 'Unreserve checked exponents'. I can't seem to do that. No biggie -- I'll eventually get to those exponents when the older work started by v24 finishes up.

2) That brings me to my second issue. As work started by v24 machines finish up, the results are reported back to the server... but they are reported UNDER THE V5 MACHINE NAME rather than as a v24 v4_computer. Several of the exponents shown above today in my assignments list have actually been completed and reported as composite:



No matter what I do, I can't seem to remove these exponents from my "assignments" list (as v4_computers), even though the result has already been turned in. Thus, as you can see, some of my "assignments" are not aging properly.

Is there any way to safely drop these completed assignments from my v4_computers assignment list, without losing credit for them?

Thanks!

Jester

First point assignments are NOT being removed from the v5 assignment list yet when they finish (maybe when George returns and fixed something). However, the results should be showing up in your results list and the Ghz days should be accumulating.

If you want to unreserve exponents that are NOT completed:
1. mark the check box by each exponent you want unreserved
2. mark the check box right beside the button labelled "Unreserve Checked Exponents"
3. click that button.

For you second issue: Are the machines that are reporting completion of v24 assignments already upgraded to v25? If so, this is normal behavior. All v24 assignments were migrated to a generic name of v24_computers but any that finish under a v25 machine report in under that new computer name.

[ADBjester]

Originally Posted by petrw1

First point assignments are NOT being removed from the v5 assignment list yet when they finish (maybe when George returns and fixed something). However, the results should be showing up in your results list and the Ghz days should be accumulating.

They are. That's the point. Results are showing up (under the v5 named machines) but aren't *leaving* the assignments list (under the v4_computers name).

If you want to unreserve exponents that are NOT completed:

I don't. I just don't want the exponents that v5 completed to continue to show up under my "Assignments" list, assigned to v4_computers.

For you second issue: Are the machines that are reporting completion of v24 assignments already upgraded to v25?

Yes.

If so, this is normal behavior. All v24 assignments were migrated to a generic name of v24_computers but any that finish under a v25 machine report in under that new computer name.

Which is fine, as I know I'll eventually get credit for them when George updates the Top Performers list. But why must they continue to appear as assignments to v4_computers under my account, once they have been reported, with no way to clear them from that list?

Jester

[sdbardwick]

Not to be patronizing, but are you sure you checked the box mentioned in step 2 of petrw1's instructions? I have been unreserving exponents assigned to v4_computers as they are completed by v25 clients using that method.

[petrw1]

Hmmm. I never considered dropping completed tests myself ... only tests I did NOT want to start.

Only tests that were assigned by v4 are NOT dropping. Tests assigned by v5 and completed in v5 are dropping. I thought maybe v4 were meant to stay in case George needs to verify they are reporting in correct. For example, TF test that started in v4 and completed in v5 are only getting credit for the v5 bit levels. They may be other anomalies.

[fes016]

Originally Posted by sdbardwick

Not to be patronizing, but are you sure you checked the box mentioned in step 2 of petrw1's instructions? I have been unreserving exponents assigned to v4_computers as they are completed by v25 clients using that method.

Originally Posted by petrw1

Hmmm. I never considered dropping completed tests myself ... only tests I did NOT want to start.

Only tests that were assigned by v4 are NOT dropping. Tests assigned by v5 and completed in v5 are dropping. I thought maybe v4 were meant to stay in case George needs to verify they are reporting in correct. For example, TF test that started in v4 and completed in v5 are only getting credit for the v5 bit levels. They may be other anomalies.

Uh oh, I've been doing what sdbardwick was doing, unassigning completed exponents under v4_computers. Is this a bad thing to do?


FES