IOS 15.0 was launched in October 2009, and I fondly remembered one of my stupidest moments. I also started ranting about IPv6-capable and IPv6-ready whitewashing.
The most popular post of October 2007 is still the tale of the three MTUs, followed by loads of OSPF posts.
Most popular posts from October 2009
- SSH RSA authentication works in IOS release 15.0M
- IOS release 15.0
- IOS fossils: Classful OSPF redistribution
- My stupid moments: Interface default route and Follow-up: Interface default route
- Report interface loss based on OSPF neighbor loss
- Fantastic DDoS protection: it’s getting worse
- DHCP client address change detector
- What went wrong: TCP lives in the dial-up world
- IOS packaging: Moore’s Law Won
- IPv6-capable or IPv6-ready: is it enough?
Most popular posts from October 2007
- The tale of the three MTUs
- OSPF router-id does not change when the interface IP address changes
- OSPF neighbors stuck in EXSTART
- GRE tunnel keepalives
- Running OSPF across a firewall
- Get creative: OSPF route filters
- You cannot start a Telnet session from Tcl shell
- Display OSPF SPF tree
- More details on OSPF route filters
- Display a summary of OSPF interfaces
Posts
Owww I like the one with default route pointing to Ethernet interface and proxy arp. Guess what ... I have seen this "trick" being used when customer had /16 allocated to one site, proxy arp enabled, and arp cache had 7000 entries during business hours. MPLS CER just had static route for that /16 pointing exactly towards customer router (switch). Now while this is an interesting way how to summarize, problem was that once in 24 hours router just "froze". As it turned out, NMS was set to send standard SNMP query to extract routing table and arp cache once in 24 hours. Router must felt very happy to send >7000 entries via SNMP. Fortunately this request can be filtered in snmp views on router.
ReplyDelete