Devices not reaching the DHCP server... bad switch gateway to blame?
-
@Dashrender said in Devices not reaching the DHCP server... bad switch gateway to blame?:
tried restarting the DHCP service - Gene said it's on the same device as the gateway, so probably not a windows box.. so might require a reboot of the whole gateway.
It's not. It's on the main AD server along with AD and DNS.
-
@Dashrender said in Devices not reaching the DHCP server... bad switch gateway to blame?:
I wonder if there is a latency issue on that specif cable?
We thought that that was the case, but appears not to be.
-
@art_of_shred said in Devices not reaching the DHCP server... bad switch gateway to blame?:
I should note that the gateway and DHCP server are one and the same.
I mispoke.. it was Art, not Gene.
-
Scope out of IPs?
First thing I would check if renewals work but new isn't
-
@Dashrender said in Devices not reaching the DHCP server... bad switch gateway to blame?:
What does STP have to do with this? Clearly the PC isn't a loopback to the switch.
by default STP on most switches puts ports in blocking mode then in fowarding mode after it checks it. RSTP does a better job of handling and fowards first then checks and blocks if need. if you don't enable rapid-pvst on most cisco switches you will have issues with DHCP once in a while but you will always have issues with PXE etc.
-
Have you restarted the DHCP server? It's hosted on a Windows box I wonder if there is some rot or an update that needs to be applied.
-
Maybe sure the Whitelist isn't enabled on the DHCP server.
-
@scottalanmiller said in Devices not reaching the DHCP server... bad switch gateway to blame?:
@Dashrender said in Devices not reaching the DHCP server... bad switch gateway to blame?:
tried restarting the DHCP service - Gene said it's on the same device as the gateway, so probably not a windows box.. so might require a reboot of the whole gateway.
It's not. It's on the main AD server along with AD and DNS.
how is a DC the default gateway?
-
Looks like split brain. That's a guess.
-
@Jason said in Devices not reaching the DHCP server... bad switch gateway to blame?:
@scottalanmiller said in Devices not reaching the DHCP server... bad switch gateway to blame?:
@Dashrender said in Devices not reaching the DHCP server... bad switch gateway to blame?:
tried restarting the DHCP service - Gene said it's on the same device as the gateway, so probably not a windows box.. so might require a reboot of the whole gateway.
It's not. It's on the main AD server along with AD and DNS.
how is a DC the default gateway?
It's not. No relationship.
-
I was having a brain fart when I said the gateway was the dhcp server. The primary dc is the server. It ended up being that it was set up to failover to a secondary dc (dhcp, that is) but the config wasn't completed on the second dc. It must have tried to failover and got stuck. It would renew leases just fine, but no new leases handed out. Removed that config and it recovered instantly. Problem solved.
-
@art_of_shred said in Devices not reaching the DHCP server... bad switch gateway to blame?:
I was having a brain fart when I said the gateway was the dhcp server. The primary dc is the server. It ended up being that it was set up to failover to a secondary dc (dhcp, that is) but the config wasn't completed on the second dc. It must have tried to failover and got stuck. It would renew leases just fine, but no new leases handed out. Removed that config and it recovered instantly. Problem solved.
It's easy enough to have them sync their configs to keep failover setup. works great for reboots and such.
-
@Jason said in Devices not reaching the DHCP server... bad switch gateway to blame?:
@art_of_shred said in Devices not reaching the DHCP server... bad switch gateway to blame?:
I was having a brain fart when I said the gateway was the dhcp server. The primary dc is the server. It ended up being that it was set up to failover to a secondary dc (dhcp, that is) but the config wasn't completed on the second dc. It must have tried to failover and got stuck. It would renew leases just fine, but no new leases handed out. Removed that config and it recovered instantly. Problem solved.
It's easy enough to have them sync their configs to keep failover setup. works great for reboots and such.
True, but it was even easier to remove the config on the dc that wasn't even in use.
-
@art_of_shred said in Devices not reaching the DHCP server... bad switch gateway to blame?:
I was having a brain fart when I said the gateway was the dhcp server. The primary dc is the server. It ended up being that it was set up to failover to a secondary dc (dhcp, that is) but the config wasn't completed on the second dc. It must have tried to failover and got stuck. It would renew leases just fine, but no new leases handed out. Removed that config and it recovered instantly. Problem solved.
Aww heard about that fail over setup, but never used it.
-
Not sure how other admins feel about this approach, but depending on the size of the environment, I like to have one DC handle all the DHCP services, keep regular dhcp backups, and have the DHCP role installed on the other DCs, but in an inactive state. If the DHCP server dies for any reason, you can easily assign another DC to handle this, do a quick DHCP restore, and you're back in business.
-
@Shuey said in Devices not reaching the DHCP server... bad switch gateway to blame?:
Not sure how other admins feel about this approach, but depending on the size of the environment, I like to have one DC handle all the DHCP services, keep regular dhcp backups, and have the DHCP role installed on the other DCs, but in an inactive state. If the DHCP server dies for any reason, you can easily assign another DC to handle this, do a quick DHCP restore, and you're back in business.
That or just install the role on another DC as needed. It's not like it takes much time. Obviously the nature of the business would determine whether you have that luxury.
-
If you have two DCs I'm not sure why you wouldn't setup DHCP failover it's a 5min setup. You already have DNS on both so why not protect from issues?
All of ours are loadbalanced/failover
