14:27:22 [illumos-gate] 17720 smatch should define __illumos__ -- Toomas Soome 14:35:51 git diff 14:35:55 oops. 21:04:45 hrm.. 21:05:09 if there's multiple equal cost routes for a destination, there's no simple way to get those with a command is there? 21:05:22 i.e. 'route get destination' is going to just pick one 21:33:32 netstat -rnv does not help? 21:33:52 well i'm thinking more something that'll do the work of matching a given destination 21:34:09 like route get does (just it only prints the one it'd pick at that time) 21:34:16 + grep/awk :D 21:34:33 i might just write something... 21:35:23 we sometimes run into issues where customers will have routes for 'data' traffic vs. admin traffic, but sometimes don't always get it right 21:35:46 so you can run into the hair pulling situation where stuff sometimes works 21:35:51 and sometimes doesn't inexplicably 21:36:02 ofc it would be nice if route get would print all.. 21:36:47 which if we can do a screen share, and the person is familiar enough with their networks, ok... but sometimes things are isolated and we can't get that 21:37:49 and trying to explain how to calculate network ranges from a netmask, etc. isn't always something a customer is going to want to do... 21:38:15 or maybe 'route get -a destination' ?? 21:38:25 (I need to see how it's doing that to see how painful that'd be) 21:51:06 Wow, there's an "ifindex_lookup" function in ip_ftable.c that doesn't seem to have a consumer. 21:51:49 Looks like leftover from pre-datapath-refactor. 21:53:44 But ire_round_robin... those also depend on ports... 21:53:56 It's the 1990s networking-team interview question: 21:54:28 "I have two equal-cost paths... how do YOU think we should select which path to use?" 21:54:53 (for multiple values of networking-team...) 22:04:13 you ask the path that always lies what his brother would say 22:05:16 dis is dae wae 22:06:04 and then one asks the disk that always stands what her sister would tell