/me/purchases times out for users with thousands of purchases

[dsas]

Quick summary

See the parent issue for more information

/me/purchases returns a HTTP 504 - gateway timeout for users with ~3000 subs. This has been noticed on /domains/manage but will probably cause them problems on other calypso pages too.

This happens for at least one user - it sounds like there have been repeated problems with this user's account for example p1694033164744309-slack-C096PD42U

Steps to reproduce

1. Switch to the user mentioned in the first message here: p1732188303637839-slack-C07GZ2UA3TN
2. Open up developer tools and switch to the network tab
3. Go to /domains/manage
4. In the network tab, filter for purchases and wait

What you expected to happen

A bunch of results

What actually happened

A http 504

Impact

One

Available workarounds?

No and the platform is unusable

If the above answer is "Yes...", outline the workaround.

No response

Platform (Simple and/or Atomic)

No response

Logs or notes

No response

[github-actions[bot]]

OpenAI suggested the following labels for this issue:

• [Feature Group] WPCOM Store & Purchases: The issue is related to the 'purchases' endpoint which suggests it's directly tied to the store and billing functionalities.
• [Feature] Purchase Management: The problem affects users with a large number of purchases, indicating issues with managing these purchases.
• [Feature] Site Performance: The HTTP 504 timeout clearly points to a performance issue when users attempt to access their purchase data.

[dsas]

@Automattic/shilling I think this is usually your department?

[sirbrillig]

Yes. This is a known issue: https://github.com/Automattic/payments-shilling/issues/2199

Many users will see an improvement when https://github.com/Automattic/payments-shilling/issues/3173 merges but it doesn't look like those performance improvements will help the specific user in the original issue. We'll need to make further improvements to that endpoint.