A two parter:
First, I need a good public time server. I found a list, tried them all, could not connect to like 30 that I tried. I was using an rdate -s somthing.arizona.edu command for a year and a half to keep four servers running with the same time. Is rdate no longer used or something? I looked into ntp programs, but I thought rdate was an ntp syncer?
Nevermind, I was acting like a moron and forgot that at one time the main server had the secondary server's IP. On boot it reactivated and conflicted.
Secondly, a scenario. Two servers are colocated at a cheap facility. Unlike my servers, they aren't hooked up to a main switch. They're on a cheapy switch or hub (probably). Both have dual Intel 10/100 NIC's, both have nearly equivelent hardware, both run updated Redhat 7.0. Both got rebooted when a boss panicked about one being down. Both come back up (after spending half an hour checking their inproperly unmounted partitions). server 1 can ping server 2 just fine, front and back (the second network cards are hooked together with a cross over for backups, server 2 is a hot spare). Server 2 can access the back of server 1, but not the front. Like it can't initiate connections. Server 1 can ssh, ping, etc on the front end to server 2, not vice versa. I do an `ifdown eth0` and `ifup eth0` on server 2 and it doesn't help. Try it on server 1 and things are fine. I'm thinking that little switch didn't have it's things straight, and needed to refresh. ARP tables, whatnot, I don't know. What are the chances it's the main server, what are the chances it's the switch/hub/cabling?
First, I need a good public time server. I found a list, tried them all, could not connect to like 30 that I tried. I was using an rdate -s somthing.arizona.edu command for a year and a half to keep four servers running with the same time. Is rdate no longer used or something? I looked into ntp programs, but I thought rdate was an ntp syncer?
Nevermind, I was acting like a moron and forgot that at one time the main server had the secondary server's IP. On boot it reactivated and conflicted.
Secondly, a scenario. Two servers are colocated at a cheap facility. Unlike my servers, they aren't hooked up to a main switch. They're on a cheapy switch or hub (probably). Both have dual Intel 10/100 NIC's, both have nearly equivelent hardware, both run updated Redhat 7.0. Both got rebooted when a boss panicked about one being down. Both come back up (after spending half an hour checking their inproperly unmounted partitions). server 1 can ping server 2 just fine, front and back (the second network cards are hooked together with a cross over for backups, server 2 is a hot spare). Server 2 can access the back of server 1, but not the front. Like it can't initiate connections. Server 1 can ssh, ping, etc on the front end to server 2, not vice versa. I do an `ifdown eth0` and `ifup eth0` on server 2 and it doesn't help. Try it on server 1 and things are fine. I'm thinking that little switch didn't have it's things straight, and needed to refresh. ARP tables, whatnot, I don't know. What are the chances it's the main server, what are the chances it's the switch/hub/cabling?
Facebook
Twitter