Envoyer Clone New Release 429 Too Many Requests Issue

Hi,

I’ve recently been working on a project that uses Laravel Forge, Laravel Envoyer and GitLab. For background Laravel Forge is a Saas product that helps you manage servers and Envoyer is a deployment tool that helps with versioning and deployments to multiple servers, handles environment variable loading and such things. The project I’ve been working on has landed an issue with the rate limit in the way Envoyer “clones” the Gitlab repository for deployment to a server. All has been working well with < 5 servers but with Covid the demand has increased and rather than increasing the server size the desire was to add a 6th server to the pool. However, this is where we discovered the 5 requests per minute restriction with Envoyer/Gitlab.

Naturally we don’t control the implementation of Envoyer but is using the endpoint https://gitlab.com/api/v4/projects/???/repository/archive with the rate limit the appropriate way to deploy, has Laravel Envoyer got the process wrong or is deploying to > 5 servers simultaneously really not done?

Appreciate any insight into whether Envoyer is doing something wrong or whether there is a setting in Gitlab we can adjust to improve deployment to > 5 servers.

Thanks in advance

Hi @ritey! I am inclined to agree that: deploying to > 5 servers simultaneously might be the issue, but it’s hard to be sure. I really know nothing about Envoyer, sorry!

I am going to reach out to the experts here at GitLab, but in the meantime I thought I would check in and see what kind of progress you have made! Talk soon!

@Linds

Appreciate the reply, I guess my query is more is using the rate limited endpoint of :

https://gitlab.com/api/v4/projects/???/repository/archive

Is the appropriate way for a 3rd party tool to fetch the repository code and deploy to servers or is there another way a tool could work?

If this is the usual way, is there any way to request a higher than 5 per minute request to that API endpoint?

Thanks

Did you ever get round this bug if so how?

I’m experiencing the same thing now deploying to 6 servers