Browse files

Advice on update_product_details as cron job and pulling from git.

  • Loading branch information...
1 parent d3e65c2 commit fab848157f5f746fff0be6fc724fb07f8398c156 @fwenzel fwenzel committed Aug 11, 2010
Showing with 10 additions and 8 deletions.
  1. +10 −8
@@ -71,11 +71,7 @@ To initialize the database, run:
./ syncdb
-And to fetch the initial product details data, run:
- ./ update_product_details
-finally, to initialize the search index, run:
+To initialize the search index, run:
./ update_index
@@ -84,7 +80,7 @@ with any web server. If you need a wsgi entry point, you can find one in
### Sites data
-The "website_issues" database can use the same SQL-database as the "default".
+The "website\_issues" database can use the same SQL-database as the "default".
It is used to load aggregate website issues (generated by clustering) from a
cron task. To initialize it, run:
@@ -95,15 +91,21 @@ To generate site data yourself without getting it pushed from metrics, run
./ generate_sites
### Cron jobs
-There are two jobs you want to run periodically:
+There are two jobs you may want to run periodically:
./ update_product_details # Mozilla Product Details update
- ./ update_index # update and rotate search index
+ ./ update_index -r # update and rotate search index
The frequency is up to you, but you probably want to run the search index
updates relatively frequently, while the product details can wait a little
+Note that updating the product details files like this will lead to "local
+changes" in your checkout. If you plan on pulling code updates from git
+periodically, you should leave ``lib/product_details_json`` untouched, but
+create a new directory somewhere else and change the setting
+``PROD_DETAILS_DIR`` accordingly.
### Mobile vs. Desktop site
We are using the [Django Sites Framework][sites] to distinguish between the
mobile site and the desktop site. The default is site ID 1 == desktop. If

0 comments on commit fab8481

Please sign in to comment.