The GNOME desktop environment includes a background task that tracks all file changes and indexes them so file-searches from the GNOME file browser can be fast. However this service does consume quite a lot of IO and CPU resources, particularly for software developers due to the way we use tools that generate large numbers of files. I don’t use desktop-search, and disabling it has made my development system far more responsive. Here are the steps needed to disable tracker (thanks primarily to this tip from site Linux Uprising):
systemctl --user mask \
tracker-store.service tracker-miner-fs.service \
tracker-miner-rss.service tracker-extract.service \
tracker reset --hard
# and check
tracker status # should report "is masked"
tracker daemon # should report "unavailable/not running"
Supposedly, within the GNOME Settings tool, going to “Search” and turning it off completely (in the window headerbar) will also disable it, but some people have found this doesn’t work.
One of the software applications I work on uses Git as version-control and has a traditional (if somewhat old-fashioned) branching strategy: a branch per feature/bugfix, an integration branch, a release-testing branch, and release branches. Different users/tools are responsible for merging between these branches so it can sometimes be rather difficult to know whether a local Git branch has been merged to any specific target.
My employer has a large Java-based application which uses Sybase ASE as its data store, and a mysterious stored procedure sp_jdbc_getprocedurecolumns (which we never call directly) is causing performance problems. It took quite a lot of work to track down why; if you’re having similar problems then see the explanation here.
I recently needed to track down some performance problems in a large (Java-based) application. My employer already uses Prometheus for monitoring, so it seemed time for me to get to grips with its concepts and instrument the application with appropriate metrics then create some matching Grafana dashboards.