We discovered that incorrect exposures were added to the new ThreePi DB since Feb. Here are the steps taken to correct this:

• remove/move bad minidvodbs from the rsync server. These are the bad ones:

+------------------------+------------+----------------+-----------------------------------------------+----------+
| label                  | label      | minidvodb_name | comment                                       | count(*) |
+------------------------+------------+----------------+-----------------------------------------------+----------+
| ThreePi.nightlyscience | ThreePi.V2 | (null)         | DM3PI i E-0 ps1_01_2340 visit 1               |       27 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.532 | DM3PI y W-13 ps1_01_5713 visit 1              |        3 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.539 | DM3PI y W+3 ps1_01_6631 visit 2               |      515 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.540 | stare3_00154 RA 20h21m StareCycle4, dec -13.8 |      503 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.541 | 130.3PI.y.QW-2 ps1_07_2124 visit 2            |      502 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.542 | 131.3PI.z.E+1 ps1_06_1313 visit 1             |      505 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.543 | 133.3PI.y.QE+1 ps1_06_0280 visit 1            |      503 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.544 | 135.3PI.z.E-0 ps1_11_0696 visit 2             |      502 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.545 | 134.3PI.z.E+2 ps1_10_0501 visit 1             |      504 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.546 | 135.3PI.y.QW+2 ps1_11_1496 visit 1            |      483 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.547 | 137.3PI.y.PE-0 ps1_11_0564 visit 1            |       18 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.562 | 139.3PI.00.BSE1.P1.i ps1_11_3686 visit 1      |       18 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.566 | 139.3PI.00.BNO2.P1.g ps1_11_0789 visit 1      |        2 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.570 | 139.3PI.00.E-5.Q2.y ps1_11_4079 visit 1       |       13 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.571 | 139.3PI.00.PS.Q1.r ps1_11_0663 visit 1        |        6 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.572 | 139.3PI.00.W-1.Q1.y ps1_10_5516 visit 1       |        4 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.576 | 140.3PI.00.ANO2.Q1.i ps1_10_1948 visit 1      |       42 | 
| ThreePi.nightlyscience | ThreePi.V2 | ThreePi.V2.577 | 138.3PI.y.QE+4 ps1_10_5705 visit 1            |        8 | 
+------------------------+------------+----------------+-----------------------------------------------+----------+

They have been moved to /data/ipp022.0/heather_minidvodbs. There are 18 corrupted out of 39 'new' minidvodbs.

• Copy the bad copy of the db to ipp005:

cd /data/ipp005.0/gpc1/catdirs
mkdir ThreePi.V2.20110614.corrupted
cd ThreePi.V2.20110614.corrupted/
rsync -rvua --progress heather@ipp004:/data/ipp004.0/gpc1/catdirs/ThreePi.V2/ .

This is in progress now.

• verify it is ok, and delete
• copy the good copy of the db from ipp006 ipp004. These include minidvodbs through ThreePi.340
• 2 ways to go here:
  • requeue up new exposures (ThreePi.V3)
  • set state of wait for the bad exposures (ThreePi.V2), and redo the bad minidvodbs
  • I do not know which is best