newpersn Posted October 9, 2014 Posted October 9, 2014 We got a problem with one of our DCs. It's taking a while to give out ip address to it computers on its vlans. It gets down to the point where the computer would get a 169 address before it would get the right address. As this causes problems with pupils trying to log in. Which means phones calls to us saying pupils can't login. The server is a vm. 1vcpu and 2GB of ram. (Same as the others DCs that are vlaned)
FN-GM Posted October 10, 2014 Posted October 10, 2014 What switches do you have? It could be down to Spanning Tree on the switches.
FN-GM Posted October 10, 2014 Posted October 10, 2014 In that case I am not sure how to check, I have only had experience with Cisco. Sorry
SHimmer45 Posted October 10, 2014 Posted October 10, 2014 it will be in the web GUI on the switch or it should be
newpersn Posted March 10, 2015 Author Posted March 10, 2015 sorry to drag this back up. But its seems to be taking longer. Is it server related? (Not giving out address quick enough. (i have added a 2nd vCPU about 6months ago)) Is it switch related? (this only happens on this vlan range) or is start of something?
Joanne Posted March 10, 2015 Posted March 10, 2015 do you have addresses free? I get the 169 addresses when I have run out on my scope. Right click scope and display statistics.
newpersn Posted March 10, 2015 Author Posted March 10, 2015 do you have addresses free? I get the 169 addresses when I have run out on my scope. Right click scope and display statistics. Total: 226 In Use: 140 (61%) Avaiable 86 (39%)
Joanne Posted March 10, 2015 Posted March 10, 2015 do you have WDS set up? See how long it takes to get an address via PXE? Also I've experienced issues with out of date LAN drivers on clients previously. Did you check your spanning tree on your switches?
newpersn Posted March 10, 2015 Author Posted March 10, 2015 do you have WDS set up? See how long it takes to get an address via PXE? Also I've experienced issues with out of date LAN drivers on clients previously. Did you check your spanning tree on your switches? WDS is hit and miss. can take a few tires to get it going. Mixed Intel,broadcom NICS. They seem to be alright everywhere else in the school (different vlans) What am i looking for with "Spanning Tree?" Should this be turned off, on or ? Thanks @Joanne
Joanne Posted March 10, 2015 Posted March 10, 2015 turn it on. It helps cut out broadcast traffic you see, which could interfere with dhcp.
Gaz Posted March 10, 2015 Posted March 10, 2015 Run some ping and tracert's across the vlans see if there are any timeouts or high ping times.
newpersn Posted August 10, 2015 Author Posted August 10, 2015 Sorry to bring this back up. I'm trying to reimage a few computer connected on this vlan range. I got 18 computers to get going. I have managed to image 12 computers though wds. (these have had a few reboots to get going but they are done and almost up and runnning) I got 7 computers that wont get an IP address. I can cmd in when its happening and watching ipconfig they are getting a 169 address and crashes out. I do an ip config after the crash out its got an ipaddress. I have done a tracert and i have got this: Tracing route to utilities1.xxx.xx [172.xx.32.3] over a maximum of 30 hops: 1 * * * Request timed out. 2 <1 ms <1 ms <1 ms utilities1.xxx.xx[172.xx.32.3] I have managed to reimage a whole IT room without problem. (another VLAN)
ITGuyWestMidlands Posted August 10, 2015 Posted August 10, 2015 Check is Portfast enabled on the switch interfaces?
IanT Posted August 10, 2015 Posted August 10, 2015 Is multicasting enabled on your network\switches? If your network is not correctly setup and getting hammered due to multicasting this can cause issues.
newpersn Posted August 10, 2015 Author Posted August 10, 2015 I will be checking the switches in question tomorrow. What settings should these be on. (3com and hp switches.) I'm guessing I would have to change the setting on the core switch down the edge cabinet
Blue_Cookeh Posted August 11, 2015 Posted August 11, 2015 (edited) I will be checking the switches in question tomorrow. What settings should these be on. (3com and hp switches.) I'm guessing I would have to change the setting on the core switch down the edge cabinet Change the setting on all switches that could potentially be receiving multicast traffic otherwise you'll flood your network and see exactly these kind of symptoms. Dumb switches wont handle mc traffic well either. Edited August 11, 2015 by Blue_Cookeh
v01d Posted August 11, 2015 Posted August 11, 2015 I will be checking the switches in question tomorrow. What settings should these be on. (3com and hp switches.) I'm guessing I would have to change the setting on the core switch down the edge cabinet With regards to end devices, make sure to set ports connected to end devices (essentially any device that isn't part of your spanning tree switch topology) to Edge Port. This will remove the delay in PCs getting an IP address on boot-up. http://h20565.www2.hp.com/hpsc/doc/public/imageServlet?DOCID=mmr_kc-0118675-5/MMA01_0131739.JPG 1
newpersn Posted August 11, 2015 Author Posted August 11, 2015 With regards to end devices, make sure to set ports connected to end devices (essentially any device that isn't part of your spanning tree switch topology) to Edge Port. This will remove the delay in PCs getting an IP address on boot-up. http://h20565.www2.hp.com/hpsc/doc/public/imageServlet?DOCID=mmr_kc-0118675-5/MMA01_0131739.JPG So if I change the setting on the port on the main switch in the cabinet this should be OK. The main switch in the cabinet has a link to a 48port switch which the room is connected on.
v01d Posted August 11, 2015 Posted August 11, 2015 (edited) So if I change the setting on the port on the main switch in the cabinet this should be OK. The main switch in the cabinet has a link to a 48port switch which the room is connected on. I'll reply again once I get to work in roughly a bit over an hour with more H3C/HP web gui pictures giving a (hopefully!) clearer breakdown of what to do, unless of course someone replies before me. EDIT: OK, I'm not sure which version of the interface you have on your switches, so I've used one of our H3C switches. If its a HP web gui I can also do one for that as well if the need is there. Firstly DON'T enable Edged Port on trunk ports connecting your switches. STEP 1 Back up your switch configuration! and perform these steps when any potential loss of workstation connectivity won't cause knuckles at the door/screaming on the phone. STEP 2 Enable MSTP on your switch if it isn't already enabled by logging into your web gui and going to Device -> MSTP -> Device Setup tab Change STP to Enable Change Mode to MSTP Click Apply STEP 3 Enable Edged Port on your switch access ports connected to desktop PCs by going to Port -> MSTP -> Port Setup tab Select all access ports connected to your desktop PCs in the selectable port diagram, making sure not to erroneously select any trunk ports Change the Protection for the selected access ports to Edged Port Click Apply Reboot a workstation connected to one of those ports and see if your IP address issues persist. There is more to MSTP, but that is a pretty basic set up for now. Essentially what setting Edged Port does for you, if you don't know, is allows the port to go straight into forwarding mode, which circumvents the 30-odd second listening period ports would otherwise observe before being able to process normal network traffic, which is a cause for workstation IP address issues on boot-up. Edited August 11, 2015 by v01d
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now