Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Service Fabric Mesh in production #379

Open
eugeneagafonov opened this issue Aug 6, 2019 · 2 comments
Open

Service Fabric Mesh in production #379

eugeneagafonov opened this issue Aug 6, 2019 · 2 comments

Comments

@eugeneagafonov
Copy link
Collaborator

We have found service fabric mesh to be able to provision our windows containers in about one minute and it's tempting to use it for scaling out our application workload. However, we have found some issues that prevent us to use service fabric mesh in production:

  1. The common provisioning time is about 1 minute, however, some replicas get stuck and their provision time is much longer. I think it is ok, but are there any guidelines for optimizing the provisioning time, what should we pay attention to?
  2. The quotas for service replicas and number of cores per application cannot be changed from the portal where the subscription quotas are. Is it possible to increase?
  3. We found that resources.azure.com does not show service fabric mesh applications' resources, and the only way to see what's going on is to use resources.azure.com/raw/ with an educated guess about what the resource url or json could look like. Is there any documentation, for example, the request format that would allow me to retrieve specific replica logs?
  4. Is there any publicly available roadmap for Service Fabric Mesh? I see a lot of unresolved issues in this repo, some of them got comments that this or that is planned for 2018, but it seems it didn't happen.
@mattrowmsft
Copy link
Contributor

  1. We are aware of the variance and there isn't much that you can do about it today. There is a "v2" stack internally which doesn't suffer from this problem. It will be leveraged by the new app model service when it is launched publicly. We precache some images to help with startup times. I'd like to make that more transparent through API's, but it is further down my list.

  2. Replica counts and cores can't be increased. I did recently lift the quota restrictions on applications.

  3. We do have an OpenAPI/Swagger spec. It is outdated but gives you some idea https://github.com/Azure/azure-rest-api-specs/tree/master/specification/servicefabricmesh/resource-manager

  4. Chacko has a youtube video that talks about what we have been working on. https://www.youtube.com/watch?v=oIdkbdlnmbw&list=PLlrxD0HtieHh73JryJJ-GWcUtrqpcg2Pb&index=2&t=0s . There will be a much larger public presence again soon. We never slowed down internally :-).

@Tim-Hodge
Copy link

@mattrowmsft Hi Matt, any update on the state of SF Mesh? Will we be seeing this come out of preview / tooling updated for VS 2019, etc?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants