Hello Martin,
I am also trying to get this working (first on a mac with dual
processors, and eventually on a cluster of macs.)
I went through the fsl_sub file and didn't see much for me to edit
other than the default email address (although i may be missing
something here.)
I saw no mention of processors, so I don't get where you are defining
processors=2.
Do you have to enter additional queues manually, and is this where
you are defining the processor number?
Sorry if this is a dumb question, but i am finding the SGE far from
trivial to setup.
Lokke
On Sep 13, 2007, at 3:13 PM, Martin Kavec wrote:
> Hi Kristen,
>
> at the moment would be pretty sure that I am running bedpostx in
> parallel. Before my processors were loaded modestly, and I could
> see only one xfibres process. Now there are two processes and the
> processors throttle on a full load. I am not familiar with sge, so
> I can not check what's in the queues. Regarding the sge
> installation on the mac, I mostly accepted default options just to
> get it simple and working as fast as possible. When the
> installation was compete I added fsl specific queues (short,
> veryshort, long, verylong), which was the most tricky part, but
> after some reading of the sge documentation I was successful.
> Except for the processors and slots, as Dave mentioned.
>
> I am going to run an analysis over night, so I'll have an idea
> about execution time in the morning.
>
> Martin
>
> On Thu, 13 Sep 2007, Lindgren, Kristen, Ann wrote:
>
>> Is there any way to check if bedpostx is actually running in parallel
>> correctly? I thought I followed the directions properly, but I'm not
>> entirely sure it's set up properly. Any help you could provide
>> would be
>> greatly appreciated. Thanks!
>>
>>
>>
>> Kristen
>>
>>
>>
>> ________________________________
>>
>> From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On
>> Behalf Of Dave Flitney
>> Sent: Thursday, September 13, 2007 8:11 AM
>> To: [log in to unmask]
>> Subject: Re: [FSL] bedpostx in parallel on MAC
>>
>>
>>
>> Okay, try "processors" and "slots" both equal to 2.
>>
>>
>>
>> "processors=1,2" seems to be interpreted as "processors=1" - this
>> needs
>> to reflect the number of cores available so that SGE can calculate
>> loads
>> correctly.
>>
>> "slots=1" means only 1 managed long.q process at a time - and you
>> would
>> like 2 presumably.
>>
>>
>>
>> On 13 Sep 2007, at 12:53, Martin Kavec wrote:
>>
>>
>>
>>
>>
>> Hi Dave,
>>
>>
>>
>> have you been able to find anything wrong in my settings? Thanks a
>> lot.
>>
>>
>>
>> Martin
>>
>>
>>
>> On Wednesday 12 September 2007 12:53:24 Martin Kavec wrote:
>>
>> Hi Dave,
>>
>>
>>
>> the outputs are below. Thanks.
>>
>>
>>
>> Martin
>>
>>
>>
>> On Wednesday 12 September 2007 12:02:31 Dave Flitney wrote:
>>
>> Could you send me the output of the following commands:
>>
>>
>>
>> qconf -sconf
>>
>>
>>
>> $ qconf -sconf
>>
>> global:
>>
>> execd_spool_dir /opt/SGE/default/spool
>>
>> mailer /usr/bin/mail
>>
>> xterm /usr/X11R6/bin/xterm
>>
>> load_sensor none
>>
>> prolog none
>>
>> epilog none
>>
>> shell_start_mode posix_compliant
>>
>> login_shells sh,ksh,csh,tcsh
>>
>> min_uid 0
>>
>> min_gid 0
>>
>> user_lists none
>>
>> xuser_lists none
>>
>> projects none
>>
>> xprojects none
>>
>> enforce_project false
>>
>> enforce_user auto
>>
>> load_report_time 00:00:40
>>
>> max_unheard 00:05:00
>>
>> reschedule_unknown 00:00:00
>>
>> loglevel log_warning
>>
>> administrator_mail [log in to unmask]
>>
>> set_token_cmd none
>>
>> pag_cmd none
>>
>> token_extend_time none
>>
>> shepherd_cmd none
>>
>> qmaster_params none
>>
>> execd_params none
>>
>> reporting_params accounting=true reporting=false \
>>
>> flush_time=00:00:15
>> joblog=false
>>
>> sharelog=00:00:00
>>
>> finished_jobs 100
>>
>> gid_range 20000-20100
>>
>> qlogin_command telnet
>>
>> qlogin_daemon /usr/libexec/telnetd
>>
>> rlogin_daemon /usr/libexec/rlogind
>>
>> max_aj_instances 2000
>>
>> max_aj_tasks 75000
>>
>> max_u_jobs 0
>>
>> max_jobs 0
>>
>> auto_user_oticket 0
>>
>> auto_user_fshare 0
>>
>> auto_user_default_project none
>>
>> auto_user_delete_time 86400
>>
>> delegated_file_staging false
>>
>> reprioritize 0
>>
>>
>>
>> qconf -sq long.q
>>
>>
>>
>> $ qconf -sq long.q
>>
>> qname long.q
>>
>> hostlist erasme52185.erasme.ulb.ac.be
>>
>> seq_no 0
>>
>> load_thresholds np_load_avg=1.75
>>
>> suspend_thresholds NONE
>>
>> nsuspend 1
>>
>> suspend_interval 00:05:00
>>
>> priority 0
>>
>> min_cpu_interval 24:00:00
>>
>> processors 1,2
>>
>> qtype BATCH INTERACTIVE
>>
>> ckpt_list NONE
>>
>> pe_list make
>>
>> rerun FALSE
>>
>> slots 1
>>
>> tmpdir /tmp
>>
>> shell /bin/sh
>>
>> prolog NONE
>>
>> epilog NONE
>>
>> shell_start_mode posix_compliant
>>
>> starter_method NONE
>>
>> suspend_method NONE
>>
>> resume_method NONE
>>
>> terminate_method NONE
>>
>> notify 00:00:60
>>
>> owner_list NONE
>>
>> user_lists NONE
>>
>> xuser_lists NONE
>>
>> subordinate_list NONE
>>
>> complex_values NONE
>>
>> projects NONE
>>
>> xprojects NONE
>>
>> calendar NONE
>>
>> initial_state default
>>
>> s_rt INFINITY
>>
>> h_rt INFINITY
>>
>> s_cpu INFINITY
>>
>> h_cpu INFINITY
>>
>> s_fsize INFINITY
>>
>> h_fsize INFINITY
>>
>> s_data INFINITY
>>
>> h_data INFINITY
>>
>> s_stack INFINITY
>>
>> h_stack INFINITY
>>
>> s_core INFINITY
>>
>> h_core INFINITY
>>
>> s_rss INFINITY
>>
>> h_rss INFINITY
>>
>> s_vmem INFINITY
>>
>> h_vmem INFINITY
>>
>>
>>
>> qconf -ss
>>
>>
>>
>> $ qconf -ss
>>
>> erasme52185.erasme.ulb.ac.be
>>
>>
>>
>> qconf -sel
>>
>>
>>
>> $ qconf -sel
>>
>> erasme52185.erasme.ulb.ac.be
>>
>>
>>
>> --
>>
>> Cheers, Dave
>>
>>
>>
>> Dave Flitney, IT Manager
>>
>> Oxford Centre for Functional MRI of the Brain
>>
>> E:[log in to unmask] W:+44-1865-222713 F:+44-1865-222717
>>
>> URL: http://www.fmrib.ox.ac.uk/~flitney
>>
>>
>>
>>
>>
>>
>>
>>
|