This repository has been archived by the owner on Nov 21, 2022. It is now read-only.
Releases: distributed-system-analysis/sarjitsu
Releases · distributed-system-analysis/sarjitsu
sarjitsu v2.0
For full details read this blog post
Release Highlights
- issues/todo list: remove docs/TODO.md (shifted to github issues)
- lib/backend: nodejs powered to python-flask powered. Also, threading for multi file uploads.
- commandline uploads: refactor vizit script to display single table
- setup.sh: various checks during installation
- 12 factor addons: graceful shutdown and restart scripts. better cleanup scripts for complete setup removal
Breaking Changes
- add index_sar as a separate lib
- config paths / contents changed
- vizit cmdline tool refactored
sarjitsu v1.0
First Release
For details, kindly refer to: http://arcolife.github.io/blog/2016/06/06/sarjitsu-a-project-on-visualizing-your-systems-activity/
Offerings
Sarjitsu offers visualizations for Operating System stats for:
- Paging
- Disks / block device activity*
- Kernel Tables [Inode Files, others]
- Task creation and system switching
- Hugepages
- I/O and Transfer rate statistics
- Swap usage / statistics
- Memory utilization / page usage by system
- CPU utilization*
- Queue: Load Averages
- Network statistics*
Note (): Features marked with asterix () are currently not directly
consumable through sarjitsu v1.0, due to Grafana 3.0.0's inability
to deal with nested documents inside elasticsearch (which is what
sarjitsu uses for data ingestion). However, there's a PR #4694 that
has been submitted to deal with the same (but hasn't been merged yet).
Software Stack
- Docker containers
- Python-Flask + uWSGI
- Nginx
- PostgreSQL
- Redis
- ElasticSearch
- Grafana
- Twitter Bootstrap
Architecture
The project is divided into 5 components:
- Grafana for visualizations
- Postgres (used for storing Grafana’s dashboards)
- Elasticsearch (used as a data source)
- API Engine in Python-Flask
- Web Server in Python-Flask (previously written in NodeJS)