Elasticsearch Tools developed at Media Cloud
- Not (yet) packaged
- May (or may not) be submitted to PyPI
A "top" program for Elasticsearch, showing the active tasks ordered by
CPU usage, in the spirit of the Un*x "top" program (original by
William LeFebvre c. 1984) and pg_top (https://pg_top.gitlab.io/), and
also the 4.3BSD "systat" program.
Requires "elasticsearch" Python package.
ESTop class can be subclassed to do local interpretation of data and
queries (see below).
Initial work by Phil Budne, funded by an NSF grant.
Media Cloud customized version of es_top (see above).
Reads Elasticsearch JSON "profile" data and collapses profiling tree
into input for https://github.com/brendangregg/FlameGraph/blob/master/flamegraph.pl
a tool for the graphical exploration of profiling data
(load the resulting .svg file into a browser).
Initial work by Phil Budne, funded by an NSF grant.
For each of the images below, clicking on them will send you to an interactive browser in the SVG file: Clicking on an element will zoom you in on that subtree. Click on a pastelled-out element at the bottom, or "zoom out" in the top left to zoom out. Typing CTRL-F allows you to search for an element by name.
totals (without node, index, shard)
breakdown by index:
breakdown by node:
breakdown by node and index:
breakdown by index and shard:
breakdown by node and index and shard:
Files in this repository were checked in under pre-commit checks with all kinds of poking and prodding. On Unix-ish systems "make install" should install a pre-commit environment, and "make lint" should run all checks.