reschedule - stopped working and won't start working

[VoidPilot]

hi

i have reschedule 1.9 working on all  of my machines and the other day, rechedule just stopped working on one of the machines. it just throws up a zero in every box on the left hand side of the front screen, yet there are 4,000 wus in the queue.   i have deleted it and re-installed, but no change, won't work.  i deleted it, rebooted, re-installed and still won't work.

any ideas would be gratefully appreciated

vp

[Ghost0210]

Hi VP,

Are the data and program directories correct in the settings tab in Reschedule?
I've had the same problem where for some reason the paths get changed.
Also are you getting any errors on the front tab in the bottom right when you run the app?

[VoidPilot]

to answer your questions:

1: re paths:

i pulled rescedule on two other boxes to cross validate the path and directories.

the main path is:  C:\Program Files\BOINC\
the directory path is:  C:\Documents and Settings\All Users\Application Data\BOINC\

all are the same..

just on the off chance that there was something odd, i deleted the information in the suspect reschedule programme and copied and pasted the information that was in one of the other non problematic boxes...no change.

2: re any errors

on the main tab, when i click on test, the six empty cells on the left each has a "zero" inserted into it and the message on the bottom right is "No reschedule needed".  The six cells on the right have no data in them, they are blank.  There is a message on the left hand side that says "0 slots occupied"

when i click run, the six empty cells on the left each has a "zero" inserted into it and the message on the bottom right is "No reschedule needed".  The six cells on the right have no data in them, they are blank.  There is a message on the left hand side that says "0 slots occupied"

3: the below is a copy of the log:

---------------------------
Reschedule version 1.9
Time: 06-04-2010 19:52:04
User testing for a reschedule
CPU tasks: 0 (0 VLAR, 0 VHAR)
GPU tasks: 0 (0 VLAR, 0 VHAR)
No reschedule needed
---------------------------
Reschedule version 1.9
Time: 06-04-2010 19:53:08
User testing for a reschedule
CPU tasks: 0 (0 VLAR, 0 VHAR)
GPU tasks: 0 (0 VLAR, 0 VHAR)
No reschedule needed
---------------------------
Reschedule version 1.9
Time: 06-04-2010 19:53:10
User testing for a reschedule
CPU tasks: 0 (0 VLAR, 0 VHAR)
GPU tasks: 0 (0 VLAR, 0 VHAR)
No reschedule needed
---------------------------
Reschedule version 1.9
Time: 06-04-2010 19:58:15
User testing for a reschedule
CPU tasks: 0 (0 VLAR, 0 VHAR)
GPU tasks: 0 (0 VLAR, 0 VHAR)
No reschedule needed
---------------------------
Reschedule version 1.9
Time: 06-04-2010 19:58:16
User testing for a reschedule
CPU tasks: 0 (0 VLAR, 0 VHAR)
GPU tasks: 0 (0 VLAR, 0 VHAR)
No reschedule needed
---------------------------
Reschedule version 1.9
Time: 06-04-2010 20:01:20
User forced a reschedule
Stopping BOINC application
BOINC application is stopped
Boinc applications
No SETI main application found
No SETI cuda application found
Starting BOINC application
BOINC application started
---------------------------
Reschedule version 1.9
Time: 06-04-2010 20:01:30
User testing for a reschedule
CPU tasks: 0 (0 VLAR, 0 VHAR)
GPU tasks: 0 (0 VLAR, 0 VHAR)
No reschedule needed
---------------------------
Reschedule version 1.9
Time: 06-04-2010 20:01:31
User testing for a reschedule
CPU tasks: 0 (0 VLAR, 0 VHAR)
GPU tasks: 0 (0 VLAR, 0 VHAR)
No reschedule needed
---------------------------
Reschedule version 1.9
Time: 06-04-2010 20:01:32
User testing for a reschedule
CPU tasks: 0 (0 VLAR, 0 VHAR)
GPU tasks: 0 (0 VLAR, 0 VHAR)
No reschedule needed
---------------------------
Reschedule version 1.9
Time: 06-04-2010 20:02:40
User forced a reschedule
Stopping BOINC application
BOINC application is stopped
Boinc applications
No SETI main application found
No SETI cuda application found
Starting BOINC application
BOINC application started
---------------------------
Reschedule version 1.9
Time: 06-04-2010 20:02:51
User forced a reschedule
Stopping BOINC application
BOINC application is stopped
Boinc applications
No SETI main application found
No SETI cuda application found
Starting BOINC application
BOINC application started

[Raistmer]

I would suppose some access rights problem. Maybe ReSchedule can't get access to BOINC's directories on NTFS. If they created under one user (BOINC master or whatever) and ReSchedule running under user account with very limited rights... Worth to check access rights maybe.

[VoidPilot]

now i'm stumped...

after my last post i waited a while and tried it agian at 21:15 hrs...clicked on the test button and it worked again...waited 2 minutes and clicked on the test button again and it did not work (back to the old problem)...waited another 2 minutes and clicked on run, it worked...

only thing that i can conclude at the moment is that it is not a problem with reschedule, but rather a problem with something else that is interfering with its operation...

thnx for the comments...

[perryjay]

This is just a guess but some people have been having problems remoting in to the newer BMs because apparently there was an update to Vista/ windows7 that somehow changed the networks from private to public. Changing it back cured their problems. Maybe that could be affecting the rescheduler too though I don't know how.

[loudbob]

Is this due to the 6.09 task being incompatible with reschedule?

[Marius]

VoidPilot: Its kind of weird, the rescheduler starts by checking some files, like boinc.exe, boinccmd.exe and %data\client_state.xml and %data\projects\setiathome.berkeley.edu. so in all cases it could find those files/paths but it was not able to read the data. A forced reschedule gives more a few more hints as it was even inpossible to find out if you're running cuda!

It could be that the client_state.xml has become corrupted, i have seen this a couple of times that boinc does not properly write the closing xml tag (or makes other mistakes in the xml layout). Unfortunately i cannot detect those .xml errors reliable and once this happends it quickly gets out of control and in the end this can lead to total corruption of the client_state.xml (forcing you to restart and redownload units). This is also exactly the reason why the rescheduler leaves a copy of the client_state.xml in the data directory so i was able to compare the old and new .xml searching for errors..


@loudbob: Rescheduler was tested with 6.08 units and its not posible to have both 6.08 and 6.09 beta units at the same time (its on the wishlist though). If you are only running 6.09 it should be working AFAICT (but i have not tested that situation).