I have two doubts here in DHCP Solaris environment.. could anyone please clarify me?
If a new client comes up and it needs an IP to be assigned by a DHCP server, how can it communicate with the DHCP server in the first case without any IP? The SUN doc says something about "BOOTP" agent but could not understand it. Can anyone enlighten me here??
What would happen if the client was shutting down and it sent a 'IP release request' (to the DHCP server which provided its IP) but the DHCP server in question was unavailable??
If the client wanted to renew its IP lease time and the DHCP server was unavailable, and after 20secs, other servers can renew its lease time... but what about IP release?? Or, does the original DHCP server retain the IP (after it comes back online) finding the allocated IP unused??
Could not find the answer after considerable Googling!!
Remember that ethernet can carry more than one kind of protocol; in the end it's just another level of frame. You can still broadcast without an IP, using your MAC address. The DHCP server can broadcast back.
Then the DHCP server wouldn't get the message, and would consider the IP to still be occupied.
Eventually it would time out, not being renewed, and be considered free again by the DHCP server, and possibly given to someone else should a request be made.