Thomas Anderson

TLDR: Made separate library call. Should not be a breaking too. --- ``` LogsPurge(repoID, pipeline int64, step ...int64) error ``` Looks like it is not many step IDs, but any...

Let's try. Add `build` label, please.

Does it work [on Windows](https://bun.sh/blog/bun-v1.0#bun-more-thing)?

> still experimental Perhaps, docs should be adjusted in case of WSL. In case of native support, I'm concerned more about > only the JavaScript runtime is supported; the package...

Got the same issue. Tailscale Linux client `v1.54.0`, Headscale `v0.22.3`. `1.46.1` works fine. Android client `1.55.47` works as intended. --- As a workaround we can rollout version `1.46.1`, approve routes...

Hi. I've not been using Tail/Headscale for a while. So, seems this is my last check before decommissioning. ``` # tailscale version 1.58.2 tailscale commit: 6a2a8457f0d62c775bfee3e350d2b34d85eda0f5 other commit: 8c7d4563e46a1856053912ac067fc5036c2d2b3e go...

**Environmental Info:** k3s version v1.28.5+k3s1 (5b2d1271) go version go1.20.12 Node(s) CPU architecture, OS, and Version: Linux debian12 6.1.0-17-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.69-1 (2023-12-30) x86_64 GNU/Linux Also checked on Oracle...

> The iptables rule that logs packets should be the right before the reject rule But they are not. There is log rule per netpol rule: ``` # iptables -L...

> show THE ACTUAL ALLOWED TRAFFIC Steps 7, 8, 9 from `Steps To Reproduce`

My understanding. Let's consider example from step 7. Packet going to 172.217.16.206, port 443, protocol TCP. Packet falls into `KUBE-NWPLCY-2YE6ZZCSRR6K3OXJ` chain. Let's follow it. 1. MARK match-set KUBE-DST-CMVHXIL52PEPFHAP dst MARK...