Help sir, my upstream sending me full BGP table I not ask for. Made VPS last month with Hetzner, everything fine until yesterday router crashed twice. I only ask default route sir, not 900k prefixes. My little box cannot handle this. Anyone know how fix? Thank you very much sir 🙏
My upstream is sending me a full table I didn't ask for
traffic worse than my packet loss
Ticket open with Hetzner. Waiting. Check your filter list. Went down again. Probably their side.
436 days. reboot is surrender.
Bro this happened to me before. Upstream template wrong. Since three days I fighting with Vultr same issue. Server down because RAM full from table. You check with them Important: ask their NOC
JokoNord said:
I only ask default route sir, not 900k prefixes
Full table is ~950k v4 + 180k v6. Filters at layer 2, scrubbing irrelevant here. 512MB RAM will choke. Ask upstream for max-prefix limit. Anycast won't help if box dies before it sees traffic.
mitigated 800Gbps before breakfast
How much RAM is the VPS
I had the same issue with a full table once, max-prefix is the fix
danfra said:
512MB RAM will choke
traffic worse than my packet loss
JokoNord said:
2GB RAM but still crash
JokoNord said:
2GB RAM but still crash
ip route show | wc -l when it happens. Hetzner NOC usually fixes template in 4-6 hours if you say "full table instead of default" in subject line.
mitigated 800Gbps before breakfast
Emre10 said:
Fighting with Vultr same issue
436 days. reboot is surrender.