NLB Cluster IP sometimes not forwarding port 3389

hello all,

for customer installed 4 remote desktop (windows server 2008 r2) servers. these servers virtual machines running on vmware vsphere hosts , configured 2 nic's.

• server 1: 172.17.30.31  172.17.30.32
• server 2: 172.17.30.33  172.17.30.34
• server 3: 172.17.30.35  172.17.30.36
• server 4: 172.17.30.37  172.17.30.38

for servers, first ip-adress configured network load balancing , second ip-adress management.

configuration of first lan-connection: nlb ticked , ipv4 ticked. have configured ip-address , subnet 255.255.0.0. configured lan-connection forward packets second nic, because on nic there default gateway configured on nic.

configuration of second nic: every option ticked, except ipv6. subnet 255.255.0.0, gateay 172.17.0.254 , dns server 172.17.30.28.

these 4 remote desktop servers configured participate in remote desktop farm , checked second ip-address of each server use ip-redirection. ip-redirection working andalso load balancing working properly, users automatically redirected correct server.

i configured (recommended microsoft articles) network load balancing. cluster ip-address 172.17.30.30. it's configured in unicast mode , port rule configured this: port 3389, tcp protocol, mode multiple, equal , affinity set none.

i have following problem: when connecting cluster ip-address rdp works directly message can't connect. cluster ip responding on ping time, when can't connect. if telnet cluster ip on port 3389 can't connect. connecting ip-address of server (the 1 nlb) can't connect (on servers). checked rdp connection settings on remote desktop servers , configured listen on alle interfaces. if do a netstat -an can see port 3389 listening on external addres 0.0.0.0 telling me listening on both interfaces. the thing can try few times , responding again. when responds or not responds random. working few hours , stops. working few minutes , stops few minutes. in eventviewer there nothing can find me resolving problem. searched internet few days now. there learned there more people having these kind of problems, these people using multicast mode and'the question in times not answerd or solved.

is there here how can guide me right direction solve problem? in 2 weeks should working well, because live environment moment users, 125 users.

i realy hope answer, getting headaches ;-)

kind regards,

jeroen bonenberg

update:

connection down again. doing telnet onport 3389 on the first en second server (on nlb nic) not working. third en fourth server accept rdp connection.

hi jeroen,

thank post.

as disucssed above point when users try connect nlb ip address ts not able login. when perform telnet fails think not configuration issue.

i recommed try multicast mode rather unicast since unicast block lan card traffic apart cluster ip address.

> please make sure nla not enabled.

> when facing issue please perform netdiag , check error messages.

> when facing issue please take tcp dump via wireshark ( http://www.wireshark.org/download.html ) , check packets getting dropped.

we wait update.

regards,

dhruv

 

 

Windows Server  >  Remote Desktop Services (Terminal Services)