Copied from Lancs CC techie forums:
Ok. Get this situation. Been using provided ip range of 10.79.24.x / 255.255.252.0 for past 5 or 6 years. Maxed out the available IP's in that range, so needed to expand to additional numbers - closer studying of scrap of paper that CLEO gave me years ago states that I can use 10.79.25.x to 10.79.27.x as well on the same subnet.
Fantastic - zillions of addresses to play with.
So, start to allocate addresses. Notice odd things happening - on furthur inspection it looks like there's problems when I use nslookup - Server not found. Googling shows that I need to add a reverse lookup zone for each IP range (before now I had no reverse lookups in there). So, add one for each range, and they start to populate from the clients.
However, still getting mucho oddness going on. I'm using XP pro SP2 clients, with a 2k server box for DNS (with a secondary DNS server to load balance). FYI, also have turned off the firewall to help with prob solving, and have also turned off hibernation/power saving stuff.
With the new XP pro clients on the 'new' IP range, I'm getting the following symptoms:
* Intermittent ping problems
* After a period of time (around 10-15 mins) if there has been no network activity then the client looses connection to the domain. Pings to the network, or browsing the network do not work, and end up giving a "access denied error". HOWEVER, if I spawn an IE window, then the ping starts back up immediatly. Voodoo wacko.gif
* sometimes get 1 way pings from client to server, but not back
* When the clients have stopped responding, if I ping from the servers, I get one reply, and then "timed out". That seems to happen EVERY time ... very odd.
After the machines have been on for a bit & they start playing up, the only way to get proper response times is to reboot the client.
I'm going through a set of Procurve switches - everything seems fine on there, no errors in the logs, likewise sod all of a pointer in the client eventlogs, nor the dns server logs.
As a direct comparison, I've got a set of identical machines (XP pro etc) sat in the room next door on the ORIGINAL IP range - and they seem fine. No timeouts, browse the network nicely fast etc etc.
So - looks like DNS, but I've poked everything I can think of & have hit a brick wall. Gone through the whole ping/ipconfig/nslookup/XP network troubleshooter and all seems fine, so I'm presuming it's DNS configuration, but I'm more than bewildered at the moment. I'm presuming it's something to do with the new IP range ... does that need to be specified somewhere???
As I said on the LancsNGFL forum. Those addresses are already used. Pay close attention to the subnet mask.
There are currently 1 users browsing this thread. (0 members and 1 guests)