Table of Contents

TUTORIAL

Video

Here the link to the video

Prerequisites

To run this tutorial you need:

  1. access to the ICARUS machines at CNAF. Instructions on how to request it can be found here.
  2. a valid x509 personal certificate and an enrollment to ICARUS vo. Instructions here

Connect to ICARUS ui

First you have to connect to bastion.cnaf.infn.it:

ssh -X <user>@bastion.cnaf.infn.it

Then, you could connect to ui01-icarus.cr.cnaf.infn.it or ui02-icarus.cr.cnaf.infn.it

ssh -X <user>@ui01-icarus.cr.cnaf.infn.it

Certificate Installation on a user interface

You have to install the x509 certificate into the user interface. To do so, first copy the certificate into the user interface with the command:

scp certs.p12 <user>@bastion.cnaf.infn.it:~

and from there again,

scp certs.p12 <user>@ui02-icarus.cr.cnaf.infn.it:~

and then follow the instructions here in the section Certificate Installation on a user interface.

Setup the ICARUS environment

Run the following script to setup the ICARUS environment:

source /cvmfs/icarus.opensciencegrid.org/products/icarus/setup_icarus.sh

Create proxy woth voms extensions

Create the proxy with the voms extensions:

voms-proxy-init --voms icarus-exp.org --valid 72:00

N.B. Be sure to have proxy with a suitable validity, everytime you submit a production job.

To check the validity of the proxy, run the command:

voms-proxy-info --all

Create your working area

Now you can create and go to your working area. For example

mkdir -p /storage/gpfs_data/icarus/local/users/tentiicarus/tutorial_20240219
cd /storage/gpfs_data/icarus/local/users/tentiicarus/tutorial_20240219 

Clone the repository

Now, you can clone the repository with the machinery for the production. A dedicated branch has been created just for this tutorial.

git clone -b tutorial_20240219 https://baltig.infn.it/icarus/prod-scripts.git

Customize the configuration

Now, you can have a look at the file variable.sh. A detailed description of its content can be found here.

cd prod-scripts
vim variables.sh

Change the variable YOUR_CUSTOM_RUN_LIST. For example

YOUR_CUSTOM_RUN_LIST="9331"

The list of the RUN2 runs is:

10054 9308 9327 9344 9366 9415 9515 9569 9596 9692 9726 9749 9764 9794 9838 9862 9895 9945 9967 10056 9309 9328 9345 9367 9433 9522 9570 9597 9693 9727 9750 9765 9796 9839 9870 9896 9949 9968 10057 9310 9329 9346 9380 9435 9531 9571 9599 9694 9728 9751 9766 9798 9840 9872 9897 9950 9969 10059 9311 9330 9347 9383 9436 9532 9580 9602 9695 9729 9752 9767 9799 9841 9873 9913 9951 9970 10060 9312 9331 9354 9384 9437 9533 9582 9610 9696 9730 9753 9768 9800 9842 9874 9914 9952 9971 10061 9314 9332 9355 9385 9438 9534 9583 9624 9699 9731 9754 9769 9801 9843 9875 9916 9953 9972 10062 9318 9333 9356 9386 9439 9558 9587 9642 9700 9732 9755 9781 9802 9844 9876 9917 9954 9974 10064 9319 9335 9357 9387 9441 9559 9588 9646 9704 9733 9756 9782 9804 9847 9886 9919 9956 9976 10065 9320 9336 9358 9388 9442 9560 9589 9648 9715 9735 9757 9783 9805 9849 9887 9921 9959 9977 10066 9321 9337 9359 9389 9444 9562 9590 9649 9716 9743 9758 9784 9806 9851 9888 9922 9960 9979 10067 9322 9338 9360 9390 9445 9563 9591 9673 9717 9744 9759 9785 9807 9852 9889 9924 9961 9981 10084 9323 9339 9361 9391 9448 9564 9592 9674 9720 9745 9760 9786 9814 9853 9890 9925 9963 9982 10085 9324 9340 9363 9392 9451 9565 9593 9688 9721 9746 9761 9788 9834 9854 9891 9926 9964 9983 10096 9325 9342 9364 9409 9513 9566 9594 9690 9723 9747 9762 9791 9835 9855 9892 9941 9965 9984 10097 9326 9343 9365 9412 9514 9568 9595 9691 9725 9748 9763 9792 9837 9860 9894 9944 9966 9986

Submit the jobs

Submit the jobs running the script submit_production.sh

./submit_production.sh

Monitor the jobs

You can check the jobs with the command:

condor_q -name sn-02.cr.cnaf.infn.it

In case something was wrong, you can debug the submission retrieving the log, err and out files with the command:

condor_transfer_data -name sn-02.cr.cnaf.infn.it <job id>

After a while the results of your submission should appear in /storage/gpfs_data/icarus/plain/data/tutorial_20240219.

Resubmission

If some job fails, you can resubmit the the job with the command:

./resubmit <RUN-NUMBER> <STREAMS>

For example,

./resubmit 9331 "BNBMAJORITY BNBMINBIAS"

CNAF monitoring webpage

CNAF provides a webpage to monitor the ICARUS job queue.