Jamil
Jamil
This is going to require https://github.com/firezone/environments/pull/33 to work as we need to ensure cluster state is not joined across deploys. To achieve blue/green, we'll need another set of instance groups...
Current method of spinning up all new nodes before destroying old ones seems to work well.
cc @thomaseizinger @AndrewDryga
Note to @AndrewDryga @bmanifold this means deploys can/will break access to the second and thereafter Resources on the same Gateway for any currently signed-in Clients if the resources are attempted...
Quitting and re-launching Firezone fixed the issue.
We can send this for each packet, so we need to batch (aggregate them). If the batch size is too high, we risk losing metrics when a gateway dies. Scalability:...
> We could write them locally into an sqlite DB and send that up to the portal every now and then? Hmm I was thinking of just batch-sending them every...
After chatting with @bmanifold it would also be helpful if we could separate these from direct vs relayed bytes so that we can also tally this on the backend per...
Good questions. Here are my thoughts so far, but we should probably hash this out much further: - Originally I was thinking (hoping) we might even adhere to a standard...
> I am not sure it is worth moving this into a background thread, I'd just do it as the packets flow through. Hm I see, so doing this accounting...