Mailing List Archives

Public Access

 		[Computer Systems Lab]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [HTCondor-users] Personal HTCondor Install - how to reduce number of slots?

I think you do have more than a single startd.  The address you got back from condor_status

doesnât match the address of the one that condor_config_val queried.   It looks like the foreign

startd is overwriting the 2 slots of the properly configured startd, and thatâs why all of the addresses are the same.

 

Try

 

  ps aux | grep -I condor_startd

 

or maybe

 

  condor_off -startd

 

If you have multiple startds, the condor_off will probably turn off only one of them, so wait a bit and then try

 

condor_status again.

 

-tj

 

From: HTCondor-users [mailto:htcondor-users-bounces@xxxxxxxxxxx] On Behalf Of Alex Churchill
Sent: Monday, February 27, 2017 11:29 AM
To: HTCondor-Users Mail List <htcondor-users@xxxxxxxxxxx>
Subject: Re: [HTCondor-users] Personal HTCondor Install - how to reduce number of slots?

 

Hi John,

 

Thank you for your response. If I run the two commands, it looks as though there is only one startd and the NUM_SLOTS should be 2:

 

mmxlabs@kubera:/etc/condor$    condor_status -af myaddress

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

<127.0.0.1:52708?addrs=127.0.0.1-52708>

mmxlabs@kubera:/etc/condor$ condor_config_val -startd -verbose -dump NUM_SLOTS

# Configuration from startd on kubera <127.0.0.1:3965?addrs=127.0.0.1-3965>

 

# Parameters with names that match NUM_SLOTS:

NUM_SLOTS = 2

 # at: /etc/condor/condor_config, line 82

 # expanded: 2

 

On 27 February 2017 at 17:24, John M Knoeller <johnkn@xxxxxxxxxxx> wrote:

Do you maybe have multiple startdâs reporting to a single collector?

 

Try

 

   condor_status -af myaddress

 

You should get the same value for all the slots if they are all coming from a single startd.

 

Or maybe something is overriding your config?

 

Try this:

 

condor_config_val -startd -verbose -dump NUM_SLOTS

 

Does it still show NUM_SLOTS=2 ?

 

-tj

 

From: HTCondor-users [mailto:htcondor-users-bounces@xxxxxxxxxxx] On Behalf Of Alex Churchill
Sent: Monday, February 27, 2017 11:11 AM
To: htcondor-users@xxxxxxxxxxx
Subject: [HTCondor-users] Personal HTCondor Install - how to reduce number of slots?

 

Hi All,

 

I have been using condor in the âpersonal installâ setting for a while on version 8.0.5, and could happily control number of slots and cpu by modifying condor_config.local. However, I have recently installed 8.4.2 on Ubuntu through apt-get and the personal setting appears to always use as many slots as cpus.

 

E.G. I have the following in condor_config.local settings:

NUM_SLOTS=2

 

For a while condor_status show:

 

mmxlabs@kubera:/etc/condor$ condor_status

Name               OpSys      Arch   State     Activity LoadAv Mem    ActvtyTime

 

slot1@kubera       LINUX      X86_64 Unclaimed Benchmar  0.000 64426  0+00:00:04

slot2@kubera       LINUX      X86_64 Unclaimed Idle      0.000 64426  0+00:00:05

                     Total Owner Claimed Unclaimed Matched Preempting Backfill

 

        X86_64/LINUX     2     0       0         2       0          0        0

 

               Total     2     0       0         2       0          0        0

 

After a little while (i.e. 3 minutes) extra slots appear:

 

mmxlabs@kubera:/etc/condor$ condor_status

Name               OpSys      Arch   State     Activity LoadAv Mem   ActvtyTime

 

slot10@kubera      LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:05

slot11@kubera      LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:06

slot12@kubera      LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:07

slot13@kubera      LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:08

slot14@kubera      LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:09

slot15@kubera      LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:10

slot17@kubera      LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:04

slot18@kubera      LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:05

slot19@kubera      LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:06

slot1@kubera       LINUX      X86_64 Unclaimed Idle      0.050 6442  0+00:14:53

slot20@kubera      LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:07

slot2@kubera       LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:05

slot3@kubera       LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:06

slot4@kubera       LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:07

slot5@kubera       LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:08

slot6@kubera       LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:09

slot7@kubera       LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:10

slot9@kubera       LINUX      X86_64 Unclaimed Idle      0.000 6442  0+00:15:04

                     Total Owner Claimed Unclaimed Matched Preempting Backfill

 

        X86_64/LINUX    18     0       0        18       0          0        0

 

               Total    18     0       0        18       0          0        0

 

Note that I have 20 CPUS on this machine.

 

Thank you for your help.

 

Regards,

 

Alex

 


_______________________________________________
HTCondor-users mailing list
To unsubscribe, send a message to htcondor-users-request@xxxxxxxxxxx with a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/htcondor-users

The archives can be found at:
https://lists.cs.wisc.edu/archive/htcondor-users/



 

--

Alex Churchill, PhD | Senior Technologist

Minimax Labs Ltd, 7 Ridgmount Street, London WC1E 7AE, UK
www.minimaxlabs.com | t: 020 7813 2014 | e: alex.churchill@xxxxxxxxxxxxxxx