yep.

Authzed Dedicated and Enterprise also allow for different permission systems too

Authzed hosted and enterprise both provide 99.999% uptime?

Enterprise is on your hardware, so the uptime is dependent upon your hardware and management (we offer support for it). Authzed Dedicated has defined SLOs. Authzed hosted is a serverless offering and has different guarantees since it is a shared offering

Do you have a commercial representative in Brazil?

Hello! We're wondering to use spicedb as authz solution for in a company. I made some research trying to find some use case with millions of resource relations in the database but i can't find it. Does anyone have an example?

are you looking for an example schema or example performance numbers or whether it is simply possible?

I'm not looking for a benchmark or something like that (would be good, but not a necessity). It's more like some real example (use case) or someone that have experience with a large number of resources (10 million+ for example).

Hello, Two Qs, 1) IS there a way to wipe out all schema and permissions in one go ? 2) How to view all permissions via zed (without having to supply resources and subjects) ? Is there a way ?

1) no. what, exactly, are you trying to accomplish? 2) you can view all data, but we don't provide a way to compute all the permissions

Sorry in my previous Qs, I meant to say relationships not permissions. 1) I was just trying to import a fresh schema.yaml, and it complained about the existing relationship referencing the outgoing schema. I had thousands of relationship, so I wished if there is a single button to do this, without specifying resource type and iterate etc.,. Eventually, ended up deleting the db and recreate it, for it to allow a fresh import. 2) How to view all relationships ? without specifying any relationships/resource specifically ?

are you running this locally? if so, you can go into the backing store and look at the rows. it's not recommended to make modifications to the backing store directly (because it would mess with revision numbers), but it can be an easier way to see things.

question of my own: any advice on getting the authzed-java client lib's TLS stuff working properly? I found the example here: https://github.com/authzed/authzed-java#initializing-a-client, but we're having a hell of a time getting the default

useTransportSecurity

to play nicely.

what issue are you encountering when trying to use it?

hmm... the more i look into this the more i think that our infra is wrong

we're trying to run gRPC connections through an AWS ALB because that's the default defined in our company's infra templates

but it looks like the Best Practice™️ for gRPC is actually to let the client be the load balancer, and all that sits between the client and server is a DNS call that tells the client where the servers are?

I am using it with 3 node k8s cluster. Eventually, I overrode it with drop and create spicedb to start with a clean slate again. Didn't capture the stderr output then, will try to simulate it again and provide it. Meanwhile, how do I acheive Q2 ? TIA

are you running your own RDS/postgres/whatever instance?

I recommend using the test-server to do simulation

its explicitly designed to be cleared between runs

I have a 3 node crdb and 3 node spicedb k8s cluster

are you testing performance or just the API semantics?

if the latter, test-server is the correct solution

okey, kinda trying both, yeah will take a note Could you please answer my Q2 part as well ? TIA

you can't view all relations, but you can use the ReadRelationships API to view all relationships under a specific definition

https://github.com/authzed/spicedb/issues/696. if you have any other ideas/suggestions on it, please add. I've started on the proto changes to support it

i'm working on getting spicedb set up in AWS infra. it seems like gRPC is designed to be aware of all downstream servers and do its own load balancing either according to xDS or to a [service config](https://github.com/grpc/grpc/blob/master/doc/service_config.md). does spiceDB implement one or both of those? in the meantime we could throw an NLB between the client and spicedb, but that seems like it foregoes the benefit of per-RPC load balancing.

Any SpiceDB node can handle an incoming request, so for incoming traffic you can set up your favorite load balancer Peer traffic (between spicedb nodes) has a couple of options. The easiest is to run on kubernetes and use the service for peer discovery (example here: https://github.com/authzed/spicedb/blob/b3a05f8dcd546910c60ad84b7cc83eb642a4f650/examples/k8s/example.yaml#L117-L118) If you're not running on kube, we do have provisional support for XDS for peer discovery (you'd use

xds://

instead of

kubernetes://

) that uses the upstream grpc xds client implementation, with docs here: https://github.com/grpc/grpc-go/blob/master/examples/features/xds/README.md