Hunk-Blade
asked on
In what environment would you use Microsoft DHCP server conflict detection
I would like to enable confict detection in our Microsoft Windows 2008 DHCP environment. The last time I had enabled this was in 2005, in a windows 2000 Domain, and back then I set it to 1.
In what circumstances is this needed? I understand the exra second it takes to do a single ping etc, and the benfits its provides. Also I understand that All clients newer than windows 2000 will attempt to do their own conflict detection before accepting an address.
We have Linux Clients, ILO ports, Windows 2000 clients, Windows XP clients and Windows 7 computers all using DHCP. Lately we have seen duplicate IP addresses on ILO ports, and on scopes that have both Linux clients and windows clients.
I have two questions:
(1a) Do Linux Clients and ILO ports perform conflict detection when they get a DHCP offer?
(1b)How does a windows client actually ping the address it has been offered before accepting ? does it "temporarily" join the same subnet under WHAT adddress before it pings? Its operating at layer 2 when getting its address, so what address does it use at layer 3as a source to actually perform the detection? This is not an issue when its renewing etc as it already has the address.
(2) Do people only enable the server side detection because they have clients on the scope that DO not do the client side detection, i.e linux / iLo etc.. I guess the server side detection just uses its normal IP.
In what circumstances is this needed? I understand the exra second it takes to do a single ping etc, and the benfits its provides. Also I understand that All clients newer than windows 2000 will attempt to do their own conflict detection before accepting an address.
We have Linux Clients, ILO ports, Windows 2000 clients, Windows XP clients and Windows 7 computers all using DHCP. Lately we have seen duplicate IP addresses on ILO ports, and on scopes that have both Linux clients and windows clients.
I have two questions:
(1a) Do Linux Clients and ILO ports perform conflict detection when they get a DHCP offer?
(1b)How does a windows client actually ping the address it has been offered before accepting ? does it "temporarily" join the same subnet under WHAT adddress before it pings? Its operating at layer 2 when getting its address, so what address does it use at layer 3as a source to actually perform the detection? This is not an issue when its renewing etc as it already has the address.
(2) Do people only enable the server side detection because they have clients on the scope that DO not do the client side detection, i.e linux / iLo etc.. I guess the server side detection just uses its normal IP.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
No one mentioned how the client side detection works, or the differences between the types of clients.
ASKER
This is the client side detection I was referring to : http://tools.ietf.org/html/rfc5227
Windows client does not do any conflict detection unless machine's own ARP entry seems getting overwritten. Then both machines get out from network.
Do you experience any DHCP-related problem on your network?
Do you experience any DHCP-related problem on your network?
ASKER
What about this, In all microsoft documentation?: DHCP client computers running Windows 2000 or Windows XP that obtain an IP address use a gratuitous ARP request to perform client-based conflict detection before completing configuration and use of a server offered IP address. If the DHCP client detects a conflict, it will send a DHCP decline message (DHCPDECLINE) to the server.
Also I just cannot find any information about the way linux clients behave at boot when using DHCP?
Yes gheist we are experiencing a few duplicate IP addresses where we have both linux and windows 200 cleints on the same subnet. The lease was set to one hour, and has now been increased.
I understand that a client will look into its arp cache whenever it receives and ARP frame to check that another host is not claiming an ip address, however I am much more interested about How linux clients behavem and how duplciates come about. For example a linux machine is off when client side detection is performed by a windows client, and When that machine gets turned back on, it tries to get the same address, and somehow does...
I am going to enable server side detection anyway, I was just curious about the process and no one seems to have a clear picture as to what happens for each type of client. Serious details.
Also I just cannot find any information about the way linux clients behave at boot when using DHCP?
Yes gheist we are experiencing a few duplicate IP addresses where we have both linux and windows 200 cleints on the same subnet. The lease was set to one hour, and has now been increased.
I understand that a client will look into its arp cache whenever it receives and ARP frame to check that another host is not claiming an ip address, however I am much more interested about How linux clients behavem and how duplciates come about. For example a linux machine is off when client side detection is performed by a windows client, and When that machine gets turned back on, it tries to get the same address, and somehow does...
I am going to enable server side detection anyway, I was just curious about the process and no one seems to have a clear picture as to what happens for each type of client. Serious details.
linux client would issue a broadcast arp request if address is free, but MS DHCP server will keep issuing same address on next request.
you also need a long lease time, because networked printers usually assume 1 day of lease even when shorter period is given by server.
linux clients also would try to acquire recently used address when not receiving any DHCP responses (by means of polite ARP)
windows clients will use same address when waking from sleep.
you also need a long lease time, because networked printers usually assume 1 day of lease even when shorter period is given by server.
linux clients also would try to acquire recently used address when not receiving any DHCP responses (by means of polite ARP)
windows clients will use same address when waking from sleep.
If you're using Windows servers for DHCP and set conflict detection to non-zero, how the client-side detection works is entirely irrelevant.
ASKER
Considering it is a best practice not to set this to a non zero value, how is understanding why client side detection is working and failing irrelevant?
By me changing this on the server, is all about a quick fix for a production environment, and I was really hoping for some clarity rather than many one sentance answers that have no detail in them.
By me changing this on the server, is all about a quick fix for a production environment, and I was really hoping for some clarity rather than many one sentance answers that have no detail in them.
Windows DHCP pings address using ICMP. It may miss omething.
Worst thing i have seen is printers from lexmark which disconnect from network on conflict and come back with previous address unless hard reset.
Worst thing i have seen is printers from lexmark which disconnect from network on conflict and come back with previous address unless hard reset.
client side detection if present will help with cases where conflicting machine does not play well with dhcp.
ASKER