Sorry for the flood of messages, but I know 2.5.5 is about to be 
released, so I want to provide all available info about bugs ASAP...

I discovered that slurmctld only crashes if I use "CPUs=2" in 
slurm.conf.  If I replace it with
"CPUs=1" or "Sockets=1 CoresPerSocket=2 ThreadsPerCore=1", everything is 
fine.

On 03/29/13 07:17, Jason Bacon wrote:
>
> Quick followup...
>
> I recompiled with -g -O0 to get more debug info:
>
> Program received signal SIGSEGV, Segmentation fault.
> [Switching to Thread 800e0aac0 (LWP 100538/slurmctld)]
> 0x00007fffff6f9418 in ?? ()
> (gdb) where
> #0  0x00007fffff6f9418 in ?? ()
> #1  0x0000000800bf616f in qsort () from /lib/libc.so.7
> #2  0x000000080131b06e in _block_sync_core_bitmap (job_ptr=0x802033a08,
>       cr_type=1) at dist_tasks.c:489
> #3  0x000000080131c2cd in cr_dist (job_ptr=0x802033a08, cr_type=1)
>       at dist_tasks.c:896
> #4  0x00000008013223f3 in cr_job_test (job_ptr=0x802033a08,
>       bitmap=0x8020413e0, min_nodes=1, max_nodes=500000, req_nodes=1,
> mode=0,
>       cr_type=1, job_node_req=NODE_CR_ONE_ROW, cr_node_cnt=1,
>       cr_part_ptr=0x800e0cd68, node_usage=0x800eac668, exc_core_bitmap=0x0)
>       at job_test.c:2530
> #5  0x0000000801316a42 in _run_now (job_ptr=0x802033a08,
> bitmap=0x8020413e0,
>       min_nodes=1, max_nodes=500000, req_nodes=1, job_node_req=1,
>       preemptee_candidates=0x0, preemptee_job_list=0x7fffff6f9a90,
>       exc_core_bitmap=0x0) at select_cons_res.c:1498
> #6  0x0000000801317e63 in select_p_job_test (job_ptr=0x802033a08,
>       bitmap=0x8020413e0, min_nodes=1, max_nodes=500000, req_nodes=1,
> mode=0,
>       preemptee_candidates=0x0, preemptee_job_list=0x7fffff6f9a90,
>       exc_core_bitmap=0x0) at select_cons_res.c:2002
> #7  0x00000000004cb6d3 in select_g_job_test ()
> #8  0x000000000046160f in deallocate_nodes ()
> #9  0x0000000000460b0d in deallocate_nodes ()
> #10 0x00000000004621cd in select_nodes ()
> ---Type<return>  to continue, or q<return>  to quit---
> #11 0x000000000043e1d8 in rehash_jobs ()
> #12 0x000000000043e66e in job_allocate ()
> #13 0x000000000046d439 in validate_operator ()
> #14 0x000000000046b978 in slurmctld_req ()
> #15 0x000000000042fec9 in main ()
>
> On 03/28/13 20:33, Jason bacon wrote:
>> I just ported slurm to FreeBSD, with very few issues.
>>
>> I've set up two simple test clusters, one on old P4 systems and one on
>> AMD64 systems.
>>
>> I have not run into any issues on the P4 systems.
>>
>> On the AMD64 systems, everything seems to work fine except srun, which
>> causes slurmctld to crash.
>> Has anyone seen something like this before?
>>
>> Thanks,
>>
>>        Jason
>>
>> Controller or compute node:
>>
>> <<<ROOT@porpoise>>>   /home/bacon 850 # sinfo
>> PARTITION         AVAIL  TIMELIMIT  NODES  STATE NODELIST
>> default-partitio*    up   infinite      1   idle porpoise
>> <<<ROOT@porpoise>>>   /home/bacon 851 # scontrol show partitions
>> PartitionName=default-partition
>>       AllocNodes=ALL AllowGroups=ALL Default=YES
>>       DefaultTime=NONE DisableRootJobs=NO GraceTime=0 Hidden=NO
>>       MaxNodes=UNLIMITED MaxTime=UNLIMITED MinNodes=1
>>       Nodes=porpoise
>>       Priority=1 RootOnly=NO ReqResv=NO Shared=NO PreemptMode=OFF
>>       State=UP TotalCPUs=2 TotalNodes=1 DefMemPerNode=UNLIMITED
>> MaxMemPerNode=UNLIMITED
>>
>> <<<ROOT@porpoise>>>   /home/bacon 852 # srun -n1 hostname
>> srun: error: slurm_receive_msg: Zero Bytes were transmitted or received
>> srun: error: Unable to allocate resources: Zero Bytes were transmitted
>> or received
>>
>> Controller:
>>
>> GNU gdb 6.1.1 [FreeBSD]
>> Copyright 2004 Free Software Foundation, Inc.
>> GDB is free software, covered by the GNU General Public License, and you are
>> welcome to change it and/or distribute copies of it under certain
>> conditions.
>> Type "show copying" to see the conditions.
>> There is absolutely no warranty for GDB.  Type "show warranty" for details.
>> This GDB was configured as "amd64-marcel-freebsd"...(no debugging
>> symbols found)...
>> (gdb) run -D
>> Starting program: /usr/local/sbin/slurmctld -D
>> [New LWP 100380]
>> [New Thread 800e041c0 (LWP 100380/initial thread)]
>> slurmctld: pidfile not locked, assuming no running daemon
>> slurmctld: Accounting storage NOT INVOKED plugin loaded
>> slurmctld: slurmctld version 2.5.4 started on cluster cluster
>> slurmctld: Munge cryptographic signature plugin loaded
>> slurmctld: Consumable Resources (CR) Node Selection plugin loaded with
>> argument 1
>> slurmctld: preempt/none loaded
>> slurmctld: Checkpoint plugin loaded: checkpoint/none
>> slurmctld: Job accounting gather NOT_INVOKED plugin loaded
>> slurmctld: debug:  No backup controller to shutdown
>> slurmctld: switch NONE plugin loaded
>> slurmctld: debug:  Reading slurm.conf file: /usr/local/etc/slurm.conf
>> slurmctld: topology NONE plugin loaded
>> slurmctld: debug:  No DownNodes
>> slurmctld: sched: Backfill scheduler plugin loaded
>> [New Thread 800e77540 (LWP 100245/slurmctld)]
>> slurmctld: Recovered state of 1 nodes
>> slurmctld: Recovered information about 0 jobs
>> slurmctld: cons_res: select_p_node_init
>> slurmctld: cons_res: preparing for 1 partitions
>> slurmctld: debug:  Updating partition uid access list
>> slurmctld: Recovered state of 0 reservations
>> slurmctld: State of 0 triggers recovered
>> slurmctld: read_slurm_conf: backup_controller not specified.
>> slurmctld: cons_res: select_p_reconfigure
>> slurmctld: cons_res: select_p_node_init
>> slurmctld: cons_res: preparing for 1 partitions
>> slurmctld: Running as primary controller
>> slurmctld: debug:  Priority BASIC plugin loaded
>> [New Thread 800e77380 (LWP 100322/slurmctld)]
>> [New Thread 800e771c0 (LWP 100361/slurmctld)]
>> [New Thread 800e0ae40 (LWP 100397/slurmctld)]
>> [New Thread 800e0ac80 (LWP 100414/slurmctld)]
>> slurmctld: debug:  power_save module disabled, SuspendTime<   0
>> [Thread 800e0ac80 (LWP 100414/slurmctld) exited]
>> slurmctld: debug:  Spawning registration agent for porpoise 1 hosts
>> [New Thread 800e0aac0 (LWP 100414/slurmctld)]
>> [New Thread 80201fc80 (LWP 100509/slurmctld)]
>> [New Thread 80201fac0 (LWP 100524/slurmctld)]
>> [New Thread 800eb9380 (LWP 100534/slurmctld)]
>> slurmctld: auth plugin for Munge (http://code.google.com/p/munge/) loaded
>> [New Thread 801c54c80 (LWP 100535/slurmctld)]
>> [Thread 800eb9380 (LWP 100534/slurmctld) exited]
>> [Thread 80201fac0 (LWP 100524/slurmctld) exited]
>> slurmctld: debug:  validate_node_specs: node porpoise registered with 0 jobs
>> [Thread 801c54c80 (LWP 100535/slurmctld) exited]
>> [Thread 80201fc80 (LWP 100509/slurmctld) exited]
>> [Thread 800e0aac0 (LWP 100414/slurmctld) exited]
>> [New Thread 800eb9380 (LWP 100414/slurmctld)]
>>
>> Program received signal SIGSEGV, Segmentation fault.
>> [Switching to Thread 800eb9380 (LWP 100414/slurmctld)]
>> 0x00007fffff3f5ee8 in ?? ()
>> (gdb) where
>> #0  0x00007fffff3f5ee8 in ?? ()
>> #1  0x0000000800b9f16f in qsort () from /lib/libc.so.7
>> #2  0x000000080131876b in _block_sync_core_bitmap ()
>>       from /usr/local/lib/slurm/select_cons_res.so
>> #3  0x00000008013198b7 in cr_dist ()
>>       from /usr/local/lib/slurm/select_cons_res.so
>> #4  0x000000080131e76b in cr_job_test ()
>>       from /usr/local/lib/slurm/select_cons_res.so
>> #5  0x0000000801316c6d in select_p_job_test ()
>>       from /usr/local/lib/slurm/select_cons_res.so
>> #6  0x00000000004536f8 in list_find_feature ()
>> #7  0x00000000004540cf in list_find_feature ()
>> #8  0x0000000000456cd5 in select_nodes ()
>> #9  0x000000000043d80f in job_allocate ()
>> #10 0x000000000045c3a0 in validate_super_user ()
>> #11 0x000000000046475d in slurmctld_req ()
>> #12 0x000000000042faa5 in main ()
>>
>


-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Jason W. Bacon
[email protected]
http://personalpages.tds.net/~jwbacon
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Reply via email to