Re: [Condor-users] discoverJobRequirements() crashing schedd

[Matthew Farrellee <matt@xxxxxxxxxxx>]

Pawel,

You're not missing anything, this was a bug in the code. It has now fixed. The fix will be available in the next developer release.

matt

On Jun 13, 2005, at 12:53 PM, Pawel.Micun@xxxxxxxxxxxxxxxx wrote:

Hello,

I'm having a problem with discoverJobRequirements() crashing schedd.

Java code using Axis:
TransactionAndStatus t = sched.beginTransaction(300);
Transaction trans = t.getTransaction();

IntAndStatus cluster = sched.newCluster(trans);
IntAndStatus job = sched.newJob(trans,cluster.getInteger());

ClassAdStructAndStatus ad = sched.createJobTemplate(cluster.getInteger(),job.getInteger(),"user",UniverseType.VANILLA,"workers.exe","argument","OPSYS==\"WINNT51\"");

RequirementsAndStatus req = sched.discoverJobRequirements(ad.getClassAd());


SchedLog:
6/13 13:41:21 Received HTTP POST connection from <1.2.3.4:1757>
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 About to serve HTTP request...
6/13 13:41:21 Completed servicing HTTP request
6/13 13:41:21 Received HTTP POST connection from <1.2.3.4:1758>
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 About to serve HTTP request...
6/13 13:41:21 SOAP leaving condor__beginTransaction() id=-1379581696
6/13 13:41:21 Completed servicing HTTP request
6/13 13:41:21 Received HTTP POST connection from <1.2.3.4:1759>
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 About to serve HTTP request...
6/13 13:41:21 SOAP leaving condor__newCluster() res=0
6/13 13:41:21 Completed servicing HTTP request
6/13 13:41:21 Received HTTP POST connection from <1.2.3.4:1760>
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 About to serve HTTP request...
6/13 13:41:21 mkdir(C:\Condor/spool\cluster1.proc0.subproc0) succeeded.
6/13 13:41:21 SOAP leaving condor__newJob() res=0
6/13 13:41:21 Completed servicing HTTP request
6/13 13:41:21 Received HTTP POST connection from <1.2.3.4:1761>
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 About to serve HTTP request...
6/13 13:41:21 Completed servicing HTTP request
6/13 13:41:21 Received HTTP POST connection from <1.2.3.4:1762>
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 Current Socket bufsize=8k
6/13 13:41:21 About to serve HTTP request...
6/13 13:41:22 entering FileTransfer::SimpleInit


core.schedd.win32:
//=====================================================
Exception code: C0000005 ACCESS_VIOLATION
Fault address:  004083EE 01:000073EE C:\Condor\bin\condor_schedd.exe

Registers:
EAX:00000000
EBX:00000000
ECX:0012D1FC
EDX:7FFB003F
ESI:0012D1FC
EDI:0090A974
CS:EIP:001B:004083EE
SS:ESP:0023:0012D1E4  EBP:0012D310
DS:0023  ES:0023  FS:003B  GS:0000
Flags:00010246

Call stack:
Address   Frame
004083EE  0012D1E4  LooseFileTransfer::getInputFiles+6
004334AC  0012D310  condor__discoverJobRequirements+75
00431545  0012D344  soap_serve_condor__discoverJobRequirements+8F
0042F5B7  0012D354  soap_serve_request+29A
0042F2EE  0012D364  soap_serve+65
0046D17F  0012FDF0  DaemonCore::HandleReq+182B
0046B782  0012FE30  DaemonCore::Driver+918
00472E07  0012FF68  dc_main+A44
00472F16  0012FF80  main+CE
00496B2D  00000001  mainCRTStartup+C5


Am I missing something?

Thanks,
Pawel