Thread: Polite complaint: network was 'best effort' for 6 hours
Posted: 2026-07-17
We've got 8x EPYC 7443 boxes at Contabo's SJC-02. Since 02:00 UTC my monitoring shows consistent 180-220ms to our upstream, normally 2ms. Ticket got "scheduled maintenance, best effort routing" response. Best effort on a 1G commit? That's 6 hours of packet loss during backup windows.
My gear:
- 4U Supermicro H12, 128GB DDR4-3200
- 2x 25G Mellanox CX-5
- 1G commit, ~300M sustained
- PDU draw 380W under load
Anyone else seeing this? I checked their status page—green across the board. EOF
visit twice: install and decom