OAR - wiki

about_the_general_data_protection_regulation

Anonymous (anonymous@undisclosed.example.com) — 2018-04-11T07:39:42+00:00

Europe is pushing for a new regulation for data protection: (translated to RGPD in french). In short, user data are sensitive data that shouldn't be conserved forever in databases. With regards to OAR, that would mean that usernames associated to jobs in a OAR database should be forgotten after some time. Another option could be to anonymize them.

about_the_job_resources_manager

Anonymous (anonymous@undisclosed.example.com) — 2021-02-11T18:22:09+00:00

Job resource manager The part of OAR that is usually responsible for setting up and cleaning up the operating system of resources given to a job is the job resource manager. This is a Perl script that is part of the OAR configuration files: located on the OAR server in

an_equivalent_to_slurm_s_srun_for_oar

Anonymous (anonymous@undisclosed.example.com) — 2017-09-14T12:55:54+00:00

Slurm's “srun” runs a given command (or script) on a reserved set of resources, ties stdout and stderr to those of the parent process, and exits with the exit code of the called script. e.g. you can do srun make It's great for batch scripts that run on the frontend.

aptly

Anonymous (anonymous@undisclosed.example.com) — 2017-07-04T20:50:00+00:00

OAR's Debian repositories setup For a list of the existing management tools for Debian packages, see: * Til August 2015, reprepro was used to manage OAR's Debian packages (thx plb). But: * it was difficult to use * it was difficult to maintain

augeas

Anonymous (anonymous@undisclosed.example.com) — 2014-11-03T14:44:35+00:00

Edit Configuration Files with Augeas From Augeas site: “Augeas is a configuration editing tool. It parses configuration files in their native formats and transforms them into a tree. Configuration changes are made by manipulating this tree and saving it back into native config files.

bug_tracking

Anonymous (anonymous@undisclosed.example.com) — 2014-12-10T13:44:59+00:00

Using github bugtracker *

coreos

Anonymous (anonymous@undisclosed.example.com) — 2014-02-28T13:16:43+00:00

CoreOS Site du projet: Un projet de gestion déploiement de serveur qui utilise systemd, docker, etcd (un outils de partage de configuration basé sur le protocole Raft) et un outils au dessus fleet comme system d'init distribué. C'est très léger et plutôt intelligent. Si vous voulez tester il y a tout ce qu'il faut ici:

coupling_oarsh_with_gnu_parallel_to_mimic_salloc_srun

Anonymous (anonymous@undisclosed.example.com) — 2020-04-16T13:45:34+00:00

Demonstration of the use of GNU parallel with oarsh, so that a user can execute batch (parallel) jobs using subsets of the resources allocated to a bigger (OAR) job. E.g. 1 batch per core (or gpu), in a job which has many nodes/cores/gpus allocated.

criu

Anonymous (anonymous@undisclosed.example.com) — 2019-03-26T16:49:52+00:00

Some work is in progress for the setup of a mechanism to allow system level checkpointing of applications using CRIU. See:

custom_admission_rules

Anonymous (anonymous@undisclosed.example.com) — 2020-03-25T14:28:17+00:00

Admission rules OAR offers a powerful system letting you customize the way that jobs enter into queues (or are rejected from queues) called “admission rules”. An admission rule is a little perl script that you insert into the admission_rules SQL table of the OAR database. Here, you'll find some advanced and useful examples.

customization_tips

Anonymous (anonymous@undisclosed.example.com) — 2020-03-25T14:24:36+00:00

Configuration In this section, you'll find advanced configuration tips Priority to the nodes with the lower workload This tip is useful for clusters of big nodes, like NUMA hosts with numerous cpus and a few nodes. When the cluster has a lot of free resources, users often wonder why their jobs are always sent to the first node while the others are completely free. With this simple trick, new jobs are sent preferably on the nodes that have the lowest 15 minutes workload.

database_slow_queries

Anonymous (anonymous@undisclosed.example.com) — 2015-07-05T16:32:01+00:00

* * * Analysers: * * pgbadger (debian pkg) -> couldn't make work * pgfouine (debian pkg) -> couldn't make work *

debian_packaging

Anonymous (anonymous@undisclosed.example.com) — 2025-02-27T12:38:54+00:00

Back to releasing Debian packaging Prepare your packaging environment See: * dput setup * sbuild-debian * aptly setup Build for sid Import origin tarball With uscan If tarball is published, use uscan: git checkout debian/2.5 # Workaround for bugs in gbp/uscan ($USCAN_DESTDIR not considered and no way to pass uscan options to the gpb import-orig --uscan the command line): uscan --no-conf -ddd --destdir=../tarballs/ gbp import-orig $(ls ../tarballs/oar_*.orig.tar.gz | tail -n1)

dedicated_command_for_queue_management

Anonymous (anonymous@undisclosed.example.com) — 2014-02-15T18:20:09+00:00

Provide a dedicated command, instead of oarnotify, and allow modifying queue instead of requiring remove+add.

developing_admission_rules

Anonymous (anonymous@undisclosed.example.com) — 2020-03-23T18:07:07+00:00

Developing admission rules The admission rules are executed within the add_micheline_job function of OAR's IO lib. See . The following perl variables can be used in admission rules: * $jobType * $ref_resource_list * $command * $infoType *

disk_reservation

Anonymous (anonymous@undisclosed.example.com) — 2016-12-02T17:33:59+00:00

This page describes a proof of concept to use OAR to couple “cpu” resources with some “disk” resources. E.g. Each node has “cpus” and “disks”, we want to allow users to create jobs to reserve disks for a long period, and eventually request smaller compute jobs on resources on which those disks are reserved. This would give users the capability of keeping data on disk for a longer time than the compute time, e.g. a more persistent storage for data.

document_the_migration_from_an_existing_mysql_database_to_postgresql

Anonymous (anonymous@undisclosed.example.com) — 2014-02-18T15:55:36+00:00

Page a creer: Migration from a Mysql to a PostgreSQL database Voir avec Pascal, migration effectuée sur Igrida.

documentation

Anonymous (anonymous@undisclosed.example.com) — 2014-02-28T11:41:17+00:00

Sphinx * 1ers essais: html : autres thèmes : Man : Epub : PDF :

dokuwiki

Anonymous (anonymous@undisclosed.example.com) — 2025-05-26T13:34:07+00:00

DokuWiki wiki:dokuwiki DokuWiki is a simple to use and highly versatile Open Source wiki software that doesn't require a database. It is loved by users for its clean and readable syntax. The ease of maintenance, backup and integration makes it an administrator's favorite. Built in

dput_setup

Anonymous (anonymous@undisclosed.example.com) — 2021-03-31T13:32:43+00:00

Obviously, we do not want to push to Debian's FTP. We push to oar-ftp using scp. oar-ftp is behind the LIG's firewall so having a LIG account is required. Also, access to the ftpmaster account must be enable (your SSH key must be added to the ftpmaster account).

editors

Anonymous (anonymous@undisclosed.example.com) — 2014-02-28T11:41:56+00:00

Sublime Text * atom.io *

enable_to_mix_timesharing_modes

Anonymous (anonymous@undisclosed.example.com) — 2016-03-02T13:26:41+00:00

Idea The idea here is to allow someone that runs a “timesharing=*,*” job, to run a new job on the resource but with the more restrictive specification: “timesharing=user,*”, if no “timesharing=*,*” jobs are running from other users. With the semantic of the Digitalis platform where:

example_of_setup_with_oar_energy_saving_mechanism

Anonymous (anonymous@undisclosed.example.com) — 2016-04-08T11:42:59+00:00

FAQ Hubot does nothing because wake-up is already in progress, but it is actually not Q: I get in the logs: [...] [Hulot] Wanted to wake up node to satisfy 'property' keepalive, but a command is already running on this node. So doing nothing and waiting for the next cycles to converge.

fix_for_database_schema_version_2.5.0

Anonymous (anonymous@undisclosed.example.com) — 2014-10-31T18:35:35+00:00

This fix applies to OAR 2.5.2 and OAR 2.5.3, and must be taken into account if upgrading to any later version of OAR (i.e. starting from OAR 2.5.4) Description of the issue With OAR 2.5.2 and OAR 2.5.3, OAR's database schema version must be 2.5.2 but is actually 2.5.0 in case of a fresh database setup with oar-database.

froggy_job_resource_manager_ht

Anonymous (anonymous@undisclosed.example.com) — 2015-07-05T08:23:18+00:00

--- /etc/oar/job_resource_manager_cgroups_memory.pl 2013-06-03 15:55:30.282952914 +0200 +++ /etc/oar/job_resource_manager_cgroups_memory_ht.pl 2014-07-07 14:36:44.730720175 +0200 @@ -32,6 +32,8 @@ my $Old_umask = sprintf("%lo",umask()); umask(oct("022")); +my $hyperthreading_type = "ht"; + my $Cgroup_mount_point = "/dev/oar_cgroups"; my $Cpuset; my $Log_level; @@ -191,6 +193,16 @@ exit_myself(5,"Failed to create and feed the cpuset $Cpuset_path_job"); } + # …

green_computing

Anonymous (anonymous@undisclosed.example.com) — 2020-03-25T14:19:31+00:00

Green computing In this page, you'll find tips for optimizing the fluids consumptions of your clusters Warning: this tip is now partly obsoleted by the new hulot module that comes with the latest oar release. this energy saving module has got a keepalive feature. take a look at the comments above all the energy* variables into the oar.conf file.

gvirt

Anonymous (anonymous@undisclosed.example.com) — 2020-03-19T12:38:19+00:00

Gvirt is a tool developed for the Igrida and Grid'5000 platform at Inria. It interfaces OAR jobs with VMs. Documentation is a work in progress. Any contribution to make it usable for other platforms would be very welcomed.

handle_systemd_api_for_managing_oar_cgroups

Anonymous (anonymous@undisclosed.example.com) — 2014-08-07T13:32:31+00:00

Apparently there are some works in progress to change the CGROUPS API in the kernel. Accordingly to they want to allow only 1 process to become a CGROUPS manager. So systemd is a good candidate. In this case we have to find a way to use the systemd CGROUPS

how_i_work_with_oar-docker

Anonymous (anonymous@undisclosed.example.com) — 2015-06-11T20:24:17+00:00

How I work with oar-docker (see also oar-docker) I'm currently using Debian 8 (jessie) as my operating system, and since I do not like to install python packages in my system outside of Debian packaging, I use the --user option of pip. But first I need to install pip from Debian

how_to_handle_a_bigger_and_bigger_oar_database

Anonymous (anonymous@undisclosed.example.com) — 2016-06-29T18:54:51+00:00

Serveral installations of OAR are known to have reached more than several millions of jobs. With the corresponding size of the OAR database, some problems may arise, first of all depending on the sizing of the database server in use. The main CIMENT cluster for instance (froggy) currently runs jobs with id > 11.000.000, with not known issues. CIMENT uses a PostreSQL database with a dedicated server (not a VM) with the follwing hardware specs:

implement_quotas_based_on_cpu-time_into_the_scheduler

Anonymous (anonymous@undisclosed.example.com) — 2014-08-04T17:07:17+00:00

Proposition (CIMENT) For a given project (or user), the jobs currently running in the system should not exceed a certain amount of cpu-time Q(project). The current cpu-time of a project T(project) is defined as the sum of all remaining cpu-times of the running jobs of a project (number_of_resources*remaining_walltime).

job_event_notifications

Anonymous (anonymous@undisclosed.example.com) — 2016-03-23T23:38:22+00:00

oarsub notification option The oarsub command provides an option to send notifications of job events. Quoting oarsub manual page, the --notify option works as follows: --notify Specify a notification method (mail or command to execute). Ex: --notify "mail:name@domain.com" --notify "exec:/path/to/script args" args are job_id,job_name,TAG,comment TAG can be: - RUNNING : when the job is launched …

job_scheduling_constraints

Anonymous (anonymous@undisclosed.example.com) — 2016-03-02T13:41:56+00:00

Allow to force jobs to be scheduled only during night for instance. Quick description Je viens de pousser le branche de travail: 2.5-sched-constraints: J'ai pas mal retravaillé le oar_meta_sched et le scheduler de base: oar_sched_timesharing_and_placeholder, ainsi que la lib GanttHoleStorage.pm.

job_types

Anonymous (anonymous@undisclosed.example.com) — 2018-10-15T15:20:21+00:00

This page aims at gathering information about the different job types (might be renamed job tags someday...), as passed to oarsub using the -t option. Some job types have effects on the job execution mechanisms, some on the job scheduling, some (user defined) can be abstract/virtual, or only have effect outsite of OAR (e.g. in the prologue/epilogue scripts set by the administrator). Job types control is managed by the admission rule. The 15th admission rule handle the syntax checking of the acc…

linux_cgroup_sysfs_and_other_sysctl

Anonymous (anonymous@undisclosed.example.com) — 2015-04-30T18:44:49+00:00

* *

managing_resources_cpu_gpu

Anonymous (anonymous@undisclosed.example.com) — 2020-03-03T13:10:04+00:00

This page presents some internals of OAR's management of resources, in order to give administrators some useful knowledge and help creating relevant definitions of the resources with regards to CPUs-cores topologies as well as CPUs-GPUs. Managing processing unit topologies

migrate_runner_module_into_almighty_bipbip_handler_processus

Anonymous (anonymous@undisclosed.example.com) — 2014-03-23T10:35:56+00:00

The goals are: * let the bipbip handler processus in Almighty to manage the launch of bipbip (for starting jobs). This will handle the window launch of processes as when the jobs are finishing. I think it is the right place. * make the starting jobs asynchronous. So when the scheduler decides that a job can be launched it is launched in background without waiting the end of the scheduler (can be > 30s).

migration_from_a_mysql_to_a_postgresql_database

Anonymous (anonymous@undisclosed.example.com) — 2015-09-23T15:11:28+00:00

Migrating a OAR database from MySQL to PostgreSQL The Migrating a OAR database from MySQL to PostgreSQL is not straight-forward. A Tool like mysql2psql does not do the trick by itself. New procedure A tool has been developed to help migrating from a mysql to a postgresql OAR database: oar-database-migrate.

multi-cluster_with_priority

Anonymous (anonymous@undisclosed.example.com) — 2021-02-11T17:21:27+00:00

For a OAR setup with different set of nodes (i.e. different homogeneous clusters) in the resources, you might want to give different priorities to nodes, so that some nodes are used in priority compared to others. For that purpose, you can use the scheduler_priority property of the nodes (setting it using the oarnodesetting command), which will be taken into account as specified per the SCHEDULER_RESOURCE_ORDER configuration setting in oar.conf.

nix

Anonymous (anonymous@undisclosed.example.com) — 2014-05-13T11:51:29+00:00

Nix Nix is a purely functional package manager. This means that it treats packages like values in purely functional programming languages such as Haskell — they are built by functions that don’t have side-effects, and they never change after they have been built.

nonexisting

Anonymous (anonymous@undisclosed.example.com) — 2013-07-10T09:34:48+00:00

This page does not exist.

oar_drmaa_interface

Anonymous (anonymous@undisclosed.example.com) — 2015-09-16T19:47:39+00:00

One can find a DRMAA interface for OAR with the project hosted at . See .

oar_skylight

Anonymous (anonymous@undisclosed.example.com) — 2017-02-22T13:55:15+00:00

Devel environment Install * Install oar-docker (follow oar3_dev_with_oar_docker) * Start and connect on frontend: oardocker start -n 2 -v ~/git/oar-skylight/:/home/docker/oar-skylight oardocker connect frontend sudo apt-get update # repeat if warnings occur sudo apt-get install -t testing libc-ares2 libssl1.0-dev sudo apt-get install -t experimental nodejs nodejs-dev sudo apt-get install npm sudo ln -s /usr/bin/nodejs /usr/local/bin/node

oar-docker-admin

Anonymous (anonymous@undisclosed.example.com) — 2017-03-06T16:24:28+00:00

Procédure de mise à jour des images oardocker sur le Docker Hub Mettre à jour l'image de debian L'image utilisée est debian stable avec quelques paquets qui viennent de testing il me semble (nginx/apache/systemd ?). Pour mettre à jour cette image :

oar-docker

Anonymous (anonymous@undisclosed.example.com) — 2020-03-31T13:43:37+00:00

Build your own OAR cluster with docker OAR-docker is a tool to setup a cluster of docker containers emulating a OAR cluster, with a frontend, a server, and possibly many nodes. OAR-docker is the perfect tool for testing OAR developments. It also allows prototyping the use of other tools on top of OAR.

oar-vagrant

Anonymous (anonymous@undisclosed.example.com) — 2024-04-10T14:20:12+00:00

Overview The oar-vagrant project aims at using Vagrant to test the different flavors of installation of OAR. For now, you can test the installation of OAR's packages (rpm, dev) on VMs running Centos 6, Centos 7, Debian 7 (wheezy) and Debian 8 (jessie), with mysql or postgresql as the database backend.

oar3_dev_with_oar_docker

Anonymous (anonymous@undisclosed.example.com) — 2017-02-03T10:20:17+00:00

Initialisation d'un environnement de dev OAR3 Nettoyage de l'environnement docker docker rm $(docker ps -a -q|head -n1) docker rm $(docker ps -a -q|head -n1) ... docker images |awk '{print "docker rmi " $3}' |bash rm -rf .oardocker Creation/chargement virtualenv

oardocker_setup_for_grid_5000

Anonymous (anonymous@undisclosed.example.com) — 2020-03-31T13:42:14+00:00

Emulating a Grid'5000 OAR setup This setup fakes the boot stages (post-deploy or standby) with at commands. It uses oar-docker. On the host Start the oar-docker cluster from within a oar git clone. * cd .../oar * oardocker -v$PWD:/mnt -n 20 On the frontend container

oarstat_vs_rest_api

Anonymous (anonymous@undisclosed.example.com) — 2016-12-12T20:40:16+00:00

OAR 2.5.7 oarstat WAITING/job batch oarstat -fj: assigned_resources= assigned_hostnames= * no information on scheduled resources oarstat -pj: empty WAITING/reservation a l'avance oarstat -fj: assigned_resources= assigned_hostnames= reserved_resources=1+2

oarwalltime

Anonymous (anonymous@undisclosed.example.com) — 2020-02-18T12:17:47+00:00

oarextratime Coming with OAR 2.5.8, oarwalltime is a new OAR command. This page talk about oarextratime, which was the former name for oarwalltime. One could just exchange the two words. Usage $ oarextratime Usage: /usr/local/lib/oar/oarextratime [] [] Manage extra time requests for a job. - If extra time is given, the command requests the addition of some extra time to the job or update a previous request. - If no extra time is given, the command shows…

pagename

Anonymous (anonymous@undisclosed.example.com) — 2013-07-10T09:34:27+00:00

Replace pagename by the name of the page you want !

passing_environment_variables_to_openmpi_nodes

Anonymous (anonymous@undisclosed.example.com) — 2024-01-30T18:44:58+00:00

Passing environment variables to nodes While some batch schedulers (e.g. Slurm) do some tricks to transparently pass some environment variables from the head node shell to all execution nodes given to mpirun, OAR does not (OAR provides no more than what OpenSSH does, be it used directly as

processors

Anonymous (anonymous@undisclosed.example.com) — 2014-02-27T16:35:58+00:00

This page aims at listing resources regarding processors (Intel, AMD, ...) and discussing enhancement for the ways of handling them in OAR... * * * * *

publication_of_the_documentation

Anonymous (anonymous@undisclosed.example.com) — 2016-02-16T07:29:44+00:00

Publishing To publish the documentation, run from OAR git: $ make -C docs publish If you don't want to publish using the the default DOC_TARGET (docmaster@oar.lig), set it as follows: $ make -C docs DOC_TARGET=my_user@my_host publish If you want to push to a specific version, set the DOC_VERSION:

releasing

Anonymous (anonymous@undisclosed.example.com) — 2015-10-26T17:18:11+00:00

This page describes the procedure to release a new version of OAR. This procedure applies to OAR version > 2.5.2 Releasing is 3 folds: * create the tarball for the release * create the new Debian packages * create the new RPM packages New Debian and/or RPM packages can also be created for an existing tarball release, to correct packaging bugs (hence, skipping step 1).

rename_oar_modules

Anonymous (anonymous@undisclosed.example.com) — 2015-05-26T14:33:39+00:00

Plan is to rename OAR module to less fancy names, e.g. oar_... We hope to only need some `sed' commands. Almighty * oarserver (-> already used as the wrapper script) bipbip * oarrunner * oarlauncher * (the runner module: does not exist anymore)

reprepro

Anonymous (anonymous@undisclosed.example.com) — 2014-10-29T22:57:42+00:00

The goal of this page is to provide useful information for installing and managing OAR's reprepro Debian repository Installation TBC Basic commands TBC Snapshoting Before copying some package version from one distribution (e.g. sid-unstable) to another (e.g. sid), it might be wise to generate a snaptshot, so that one can still configure a sources.list file providing access to an earlier version of OAR's packages.

rest-api-specs

Anonymous (anonymous@undisclosed.example.com) — 2016-11-06T22:16:45+00:00

rpm_packaging

Anonymous (anonymous@undisclosed.example.com) — 2021-01-22T14:26:40+00:00

Back to releasing RPM packaging Adapt the spec file Make sure to update the version (top of the file) and the changelog (bottom of the file) at least. workstation$ git checkout centos7/2.5 workstation$ git pull --rebase workstation$ vi rpm/SPECS/oar.spec workstation$ git add rpm/SPECS/oar.spec && git commit

running_the_api_tests

Anonymous (anonymous@undisclosed.example.com) — 2017-06-30T14:49:09+00:00

Manual testing with oardocker Set user: APIUSER=docker:docker or to get access to privileged commands APIUSER=oar:oar Then set the URL API: * from outside oardocker containers OARAPI=http://$APIUSER@localhost:46668/oarapi-priv * from inside the oardocker frontend container

sbuild-debian

Anonymous (anonymous@undisclosed.example.com) — 2020-09-28T13:30:25+00:00

This setup was guided by the following page: (Since then, that Debian wiki page evolved, update my be need on our side...) Install sbuild + initial setup apt-get install sbuild user=neyron sbuild-adduser $user # this needs "entropy"... Do weird things and think different, waiting for it to complete... sbuild-update --keygen # share the deb archives among schroots... echo /var/cache/apt/archives /var/cache/apt/archives none rw,bind 0 0 >>/etc/schroo…

schroot-centos

Anonymous (anonymous@undisclosed.example.com) — 2013-07-09T08:52:08+00:00

Setup a CentOS schroot on Debian Inspired by the setup of Debian chroot for sbuild: * * $ apt-get install schroot sbuild btrfs-tools We will use the btrfs flavor of schroot. Since btrfs is still experimental, I choose to create loopback device formated with btrfs and mounted as follows:

setup_a_oar_cluster_on_a_single_machine

Anonymous (anonymous@undisclosed.example.com) — 2014-03-25T19:30:02+00:00

Setup the machine We setup a VM with a basic installation of Debian Stable (Wheezy). We call the VM “oar” and setup root's ssh keys. Create OAR tarball and push it to the VM neyron@takelma:~$ cd scm/OAR/oar neyron@takelma:~/scm/OAR/oar (2.5 $=)$ git pull --rebase Current branch 2.5 is up to date. neyron@takelma:~/scm/OAR/oar (2.5 $=)$ make tarball ./misc/make_tarball Using the Git repository '/home/neyron/scm/OAR/oar' Using the Git tree-ish '2.5' ../tarballs/oar-2.5.3+94.g6396d49.tar.gz n…

some_examples_of_admission_rules

Anonymous (anonymous@undisclosed.example.com) — 2020-04-07T20:01:29+00:00

Admission rules are managed with the oaradmissionrule command. See the usage or man page of the command. The code of the admission rules is executed within oarsub at submission time. Many internal variables are available, making the mechanism very powerful

source_control

Anonymous (anonymous@undisclosed.example.com) — 2014-02-28T11:38:31+00:00

Git * * * * * GitHub *

start

Anonymous (anonymous@undisclosed.example.com) — 2020-03-23T17:43:47+00:00

This wiki section is open to contributions from anyone (regular users, developers, ...) who's registered a login account. Developer's head pages OAR's Github page: * Todos * Testing * Releasing * Oar-docker-admin * Technology watch * Job types Restricted area The wiki restricted area is only accessible to the core team members. Old wiki content

syntax

Anonymous (anonymous@undisclosed.example.com) — 2025-05-26T13:34:07+00:00

Formatting Syntax DokuWiki supports some simple markup language, which tries to make the datafiles to be as readable as possible. This page contains all possible syntax you may use when editing the pages. Simply have a look at the source of this page by pressing

systemd_for_administrators

Anonymous (anonymous@undisclosed.example.com) — 2014-11-03T13:25:23+00:00

* * * * * * * * …

tarball_release

Anonymous (anonymous@undisclosed.example.com) — 2016-02-15T22:20:30+00:00

This page describe how new release tarball are created from the Git repository. Please have a look at the releasing for the overall process. If one only wants to publish a snapshot tarball, one can skip the first step, and use the snapshot branch on the archive server.

technology_watch

Anonymous (anonymous@undisclosed.example.com) — 2017-01-19T10:06:07+00:00

This page aims a listing new technologies we are watching for possible later (or current) use in OAR or OAR ecosystem. Sub-pages should provide: * quick description applied to our ecosystem * bookmarks, URLs * any example * anything else *

testing

Anonymous (anonymous@undisclosed.example.com) — 2018-07-18T11:18:34+00:00

Several subprojects aim at providing environments to test OAR: * OAR-Docker -> for OAR developers, or advanced users (see also how_i_work_with_oar-docker or oar3_dev_with_oar_docker) * OAR-Vagrant -> for OAR packaging developers and OAR users (admins who want to test with OAR installation, users who want to test a basic cluster setup with resources and tasks managed by OAR)

ui_toolkit

Anonymous (anonymous@undisclosed.example.com) — 2014-02-28T13:12:13+00:00

Kivy * External Link * Kivy: a Crossplatform Framework for NUI in Pytonh

use_a_cache_system_in_the_scheduling_algorithms

Anonymous (anonymous@undisclosed.example.com) — 2014-07-04T13:11:58+00:00

Move some gantt prediction related data from DB to cache system (like memcached or redis) OR Put the volatile gantt tables in /dev/shm for example DROP TABLESPACE IF EXISTS oarramts; CREATE TABLESPACE oarramts LOCATION '/dev/shm/oar'; CREATE UNLOGGED TABLE foo(i int) TABLESPACE oarramts;

use_cases_and_user_tips

Anonymous (anonymous@undisclosed.example.com) — 2021-07-28T13:59:52+00:00

Use cases OpenMPI + affinity We saw that the Linux kernel seems to be incapable of using correctly all the CPUs from the cpusets. Indeed, reserving 2 out of 8 cores on a node and running a code that uses 2 process, these 2 process where not well assigned to each cpu. We had to give the CPU MAP to OpenMPI to do cpu_affinity:

useful_commands_and_administration_tasks

Anonymous (anonymous@undisclosed.example.com) — 2020-03-25T14:15:07+00:00

Useful commands and administration tasks Here, you'll find useful commands, sometimes a bit tricky, to put into your scripts or administration tasks List suspected nodes without running jobs You may need this list of nodes if you want to automatically reboot them because you don't know why they have been suspected and you think that it is a simple way to clean things:

using_hyperthreading_on_nodes

Anonymous (anonymous@undisclosed.example.com) — 2015-09-04T15:24:00+00:00

On using (hyper)threads as the leafs of the resources tree If HyperThreading is enabled in BIOS, using threads as the leafs of OAR (implicit) resources tree is quite simple. Using the oar_resource_init script The last version of the oar_resource_init

using_oar_on_debian_based_systems

Anonymous (anonymous@undisclosed.example.com) — 2020-03-31T13:29:44+00:00

Deploying OAR on Debian based systems OAR is packaged in Debian, see Download. Examples See oar-vagrant: provision scripts in the deb_debian* directories give a good overview of how to perform a basic cluster installation using OAR on a Debian system.

using_oar_on_redhat_based_systems

Anonymous (anonymous@undisclosed.example.com) — 2020-03-31T13:31:58+00:00

Deploying OAR on Redhat based systems The Download page provides RPM packages. They are currently out-dated, but up-to-date packages could be built from sources. Contributions on that matter are very welcomed. Examples See oar-vagrant: provision scripts in the rpm_centos* directories give a good overview of how to perform a basic cluster installation using OAR on a Redhat like system.

web_ui

Anonymous (anonymous@undisclosed.example.com) — 2014-02-28T13:23:19+00:00

w2ui * w2ui : a javascript UI based on Jquery * * Web Admin-Template * Bootstrap-Admin-Template

welcome

Anonymous (anonymous@undisclosed.example.com) — 2025-05-26T13:34:07+00:00

Welcome to your new DokuWiki Congratulations, your wiki is now up and running. Here are a few more tips to get you started. Enjoy your work with DokuWiki, -- the developers Create your first pages Your wiki needs to have a start page. As long as it doesn't exist, this link will be red: