mirror of
https://github.com/MHSanaei/3x-ui.git
synced 2026-09-22 00:24:12 +00:00
6aa87f4e57
Delete aborted its per-inbound loop on the first error, so a client attached to inbounds across several nodes lost at most one per attempt: the loop never reached the remaining nodes, the record cleanup after the loop never ran, and each retry started over with whatever was left. Operators with many nodes had to delete the same client once per node. Collect per-inbound failures and keep going so every reachable inbound and node is cleaned in a single pass, then keep the client record only when something failed - its settings JSON still holds the client there, so the next delete retries exactly the leftovers - and return the joined failures instead of silently reporting success. DeleteByEmail's legacy fallback loop gets the same treatment. Closes #5845 |
||
|---|---|---|
| .. | ||
| config | ||
| database | ||
| eventbus | ||
| logger | ||
| mtproto | ||
| sub | ||
| tunnelmonitor | ||
| util | ||
| web | ||
| xray | ||