Advanced Aspects of the Django Ecosystem: Haystack, Celery & Fabric
summary
Three concrete tools and how they are used: Haystack connects Django models to Whoosh, Solr or ZAPEN backends so a working full‑text search over mixed models is built in hours (museum demo won a contest) and scales via Solr faceting, batching updates (100 docs) and Solr master/slave replication routed through nginx (four slaves in production). Celery runs background jobs (URL enrichment via oEmbed + HEAD/partial fetch, PDF detection), uses Redis as broker and memcached for transient results, and implements fan‑out activity streams asynchronously. Fabric automates SSH deployments, bootstraps Redis, manages symlinked releases with rollback and generates server/nginx configs from a servers.json roles file.
AI-written summary · about
chapters (18)
- 0:00Intro
- 2:25Haystack
- 7:08Pick your backend
- 15:34Staying fresh
- 19:31nginx + Solr replication trick
- 26:16Tasks?
- 27:04Trivial example
- 29:37The task itself...
- 31:04Behind the scenes...
- 33:53Configuration
- 35:44Advanced Celery
- 36:59The Activity Stream pattern
- 43:16Automated deployments
- 44:20Simple examples
- 49:16Rollback with symlinks
- 50:34servers.json
- 51:46In the fabfile.py
- 53:01Dynamic nginx config
related talks
41mProduction Django: Building a Highly Scalable, Secure Django Site
29mSwiss Army Django: Small Footprint ETL
24mLightning Talks
28mLightning talks
19mSwiss Army Django: Small Footprint ETL
47mHigh Performance Django at Ten: Old Tricks & New Picks
28mBoosting Python Web Apps performance with Celery and Concurrency
26mWhat Django Deployment is Really About