My favorites | Sign in
Project Home Downloads Wiki Issues Source
Search
for
Cron  
This page explains how does CirruxCache cron work.
Phase-Deploy, Phase-Support
Updated May 18, 2010 by sam.alba

Introduction

CirruxCache uses a garbage collecting cron task to flush expired entries.

Details

When CirruxCache make a request through the a Cache PoP, the fetched object is stored with a TTL. Once the TTL is reached, if the object is fetched again, its meta information are refreshed, with its content if needed.

But a lot of objects remain on the Datastore without being refreshed. For instance, if the resource is removed from the origin.

That's why a garbage collector needs to be called from a cron task.

Keep in mind that the delay of the garbage collector iteration influences your resource quota usage.

For instance, I use a refresh delay of:

  • 1 week on my personal blog (because static almost never change).
  • 15 minutes on big production servers (to keep storage as clean as possible).

To change the garbage collector delay, edit the file "cron.yaml" and change the "schedule" property.


Sign in to add a comment
Powered by Google Project Hosting