Skip to main content

Linter for Slurm configuration files

Project description

Travis CI Status python 3.5, 3.6, 3.7

A Simple linter for Slurm Config Files

Purpose

In my experience it is very easy to make small errors in slurm configuration files, incorrectly listing nodes in the cluster, or failing to assign nodes to partitions, and so on. This package provides a simple command to check slurm files that is suitable for use in automated deployment management.

In addition, a python API is provided allowing checking arbitrary strings of text. Utility functions such as explanding slurm-style hostname lists are also given.

Usage

The slurmlint command takes a slurm configuration file as an argument and returns a simple report with errors (if any). If there were errors then the return code is 1, otherwise 0.

$ slurmlint slurm.conf
753 compute nodes configured

Errors detected:
Line 126 - Duplicate node definition: ng1031, ng908, ng909, ...
Line 127 - Duplicate node definition: ng1031
Line 128 - Duplicate node definition: ng908, ng909, ng910
Line 131 - Defined node has no partition: ng1259
Line 131 - Duplicate node definition: ng1256

Limitations

This utility does not fully understand the grammar of slurm configuration files and does not guarantee the validity of a configuration file. It only serves to help find certain kinds of simple errors.

Hostname Lists

This linter applies stricter rules than Slurm for lists of hostnames. Hostnames must be valid hostnames according to RFC-1123, that is they may contain a-z, 0-9, dot, and the minus sign, at most 253 ASCII characters, and must be comprised of "labels" from 1-63 characters connected by dots.

Hostnames can be separated by commas in a list.

The standard (???) rules for brackets apply. If a list entry uses brackets to denote some numerical range or comma separated list, there must be brackets at the end of the hostname. One can also have a bracketed range somewhere else in the hostname provided that there are brackets at the end. Brackets may contain valid non-negative integers separated by commas, or ranges such as 1-8. A leading zero in a range denotes a fixed width, i.e. 001-020.

Note that this linter does not use the slurm parser and will result in "false positive" warnings where allowed hostname ranges still result in warnings.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date
slurmlint-0.0.2-py3-none-any.whl (9.3 kB) Copy SHA256 hash SHA256 Wheel py3
slurmlint-0.0.2.tar.gz (8.1 kB) Copy SHA256 hash SHA256 Source None

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page