Problem/Motivation

It appears that Pantheon Edge may rate limit invalidation requests in an undocumented fashion. This issue seeks to identify any rate limits and to document them. (Pantheon Support wasn't able to speak to this and recommended I open a ticket in this module's issue queue.)

Steps to reproduce

My client has a jobs board website, which updates its jobs from an external service 4 times per day. These updates includes several hundred entities, which are updated in bulk over the course of 3-4 minutes (with two update processes happening in parallel). Upon processing, each entity is saved (i.e. $entity->save()), which triggers cache invalidation.

There is an aggregation block on the front page (a view). Following a bulk entity update, the front page external cache is not invalidated; however, when a listed entity is updated manually, external cache invalidation for the front page works as expected. I have verified the proper cache tags are being bubbled up.

It strikes me this could be a rate limit issue.

Proposed resolution

Almost all APIs have rate limits. Let's document any that exist here.

Remaining tasks

Document any rate limits relating to Pantheon Edge.

User interface changes

N/A

API changes

N/A

Data model changes

N/A

Comments

Chris Burge created an issue. See original summary.

chris burge’s picture

Status: Active » Closed (works as designed)

It turned out to be an edge case in custom code.