As we believe in automation and using computers for what they do best (i.e. take care of repetitive tedious tasks for us), we have started to develop a support automation tool for NMIS.
The support tool is already part of NMIS version 8.5G, and it will incrementally grow in functionality; the first version provides a convenient way to collect all relevant info regarding an NMIS server's status and configuration for troubleshooting purposes.
The newest version of this tool can be downloaded here: (NMIS 8.5G and newer ship with the support tool preinstalled.) It needs to be saved in the
admin/ directory of your NMIS installation; most commonly that would be
Running the support tool without arguments shows a help screen.
Selftest and Automatic Repair
Starting with version 1.4.0 the Support Tool also runs the NMIS selftest, displays any problems detected - and offers to run NMIS' automatic repair procedure if correctables are detected (i.e. file ownerships, permissions and directory existence are verified and fixed if incorrect).
Collecting Support Information
When given the argument
action=collect, the script collects diagnostic and configuration information for both the host and the NMIS installation and save that in a (size-limited) ZIP file. If the tool detects installed commercial Opmantek products with the Opmantek Support Tool present, then that tool will also be run and you will be presented with two separate ZIP files.
Please note that the information collection does not cover node-specific files by default (to keep the zip file small); if you want to also collect node-specifics for one or more nodes add the argument
node='*' to cover all nodes) to the tool invocation.
The support tool limits the size of the resulting zip file to 10 megabytes, if necessary by truncating log files. If this does not reduce the size sufficiently the support tool will report that it cannot meet the target zip size and abort; In this case you should rerun the support tool with a larger maximum for the zip file (e.g.
maxzipsize=31457280 for 30 megabytes). This is not a common occurrence and you should investigate why such overly large amounts of data were present in logs and configuration.