XAMPPRocky
XAMPPRocky
We've noticed that quilkin will sometimes encounter a provider failure and seemingly recover but the readiness of the instance does not change.
Either in this, or as a followup, I think we should remove the agent distinction entirely. I think instead of just agents forwarding information about themselves, all instances should be...
Currently we make a delineation that the service configuration is only configured through the CLI and env, and this means that service configuration is available through the configuration, which both...
The submodules currently used are a pain to work with and a bit of maintenance we don't need, so we should remove them.
Token based routing currently relies on an active up to date connection to a config provider, which can be disrupted. To enable placing the routing information directly in the packet,...
This issue seems to show up between gRPC services and clients over real world network, but we haven't been able to replicate locally or in CI. ``` {"timestamp":"2025-03-12T11:09:34.635440Z","level":"ERROR","fields":{"message":"error receiving delta...
In order to enable us to have multiple replicas of a "relay" service, we need to enable replicas to share the live configuration together. Since our data is live and...
Looking at the `gh-pages` branch, the last push was a month ago, there haven't been any docs pushes since. https://github.com/googleforgames/quilkin/tree/gh-pages
Similar to #1185 we should also be including the latency data from an instance in `Config` so that the latency data is shared between instances and every instance is able...
Currently we only ping datacenters with active streams, but we should also be measuring our distance to xDS clients/proxies so that an instance that sits between datacenters and proxies has...