Win 2003: Unable to connect to cluster via NLB manager

I've configured NLB on two Win 2k3 servers. These servers have two NIC's. NLB
is configured on the front end lan (EFL) only. From a standalone server which
has 'visibility' to this cluster - I'm able to ping both the individual
hosts, ping the EFL Cluster IP (by cluster name and IP). I ran a netmon trace
and it indicated that my packets are being answered to. However, when I try
to connect to this cluster (by name or IP) from the NLB manager of the
standalone server (I mentioned above) I get the message that "NLB is not
loacted on the specified computer". I have an application that needs to
connect via this "application NLB cluster" to login to the DB - which does
not go through. I'm beginning to believe that this has something to do with
the above issue.

Any pointers on how to solve this would be greatly appreciated.

Thank you.

Hello,

I think you have two issues
-Access the NLB from the MMC
-access to a specific port on the NLB IP.

As the NLB IP is binded and reachable, i think you forgot to add the needed
TCP port on the NLB, like the 1433 for SQL.

On one node, click to add existing nlb. Type 127.0.0.1 as server to join.
Does it work ?

Why are you speaking of two network card ?
Does the DB bind on all interfaces (netstat -an will help) ?


--
Cordialement,
Mathieu CHATEAU
http://lordoftheping.blogspot.com


"Condor" <(E-Mail Removed)> wrote in message
news:E52412A0-6A05-4371-A077-(E-Mail Removed)...
> I've configured NLB on two Win 2k3 servers. These servers have two NIC's.
> NLB
> is configured on the front end lan (EFL) only. From a standalone server
> which
> has 'visibility' to this cluster - I'm able to ping both the individual
> hosts, ping the EFL Cluster IP (by cluster name and IP). I ran a netmon
> trace
> and it indicated that my packets are being answered to. However, when I
> try
> to connect to this cluster (by name or IP) from the NLB manager of the
> standalone server (I mentioned above) I get the message that "NLB is not
> loacted on the specified computer". I have an application that needs to
> connect via this "application NLB cluster" to login to the DB - which does
> not go through. I'm beginning to believe that this has something to do
> with
> the above issue.
>
> Any pointers on how to solve this would be greatly appreciated.
>
> Thank you.

Thanks Mat. I checked and the ports which my app needs are 80, 443 and 135 -
these are already added to NLB config. NLB with 127.0.0.1 works good...

I have an entry in the hosts file as well. 'UATRTCAPP' is the name of the
cluster.

10.253.152.73 uatrtcapp

I'm not a pro at this stuff but when I meant two network cards - I meant
each node is on 2 LANS - front end (EFL) and back end (EBL). Clustering is
done only on the EFL. Not sure how relevant it is to the current issue,
though.

Another thing is I can ping each node (from 3rd server) in the cluster and
'UATRTCAPP' from the third server which has visibility to the cluster. But
when I try to connect to 'UATRTCAPP' from NLB manager I always get 'Could not
locate NLB on specified computer'.

One more interesting thing is when I reboot the third 'server' and
immediately try to log into my app, it goes through. This is because NLB
manager from the third server identifies 'UATRTCAPP' But after some time,
login fails and NLB manager now gives 'Could not locate NLB.......' - same
error as above. Looks like it is intermittent. I'm currently digging thru the
Event viewer to see if I can find any errors.

Hope my explanation is not confusing. Any pointers are greatly appreciated.

"Mathieu CHATEAU" wrote:

> Hello,
>
> I think you have two issues
> -Access the NLB from the MMC
> -access to a specific port on the NLB IP.
>
> As the NLB IP is binded and reachable, i think you forgot to add the needed
> TCP port on the NLB, like the 1433 for SQL.
>
> On one node, click to add existing nlb. Type 127.0.0.1 as server to join.
> Does it work ?
>
> Why are you speaking of two network card ?
> Does the DB bind on all interfaces (netstat -an will help) ?
>
>
> --
> Cordialement,
> Mathieu CHATEAU
> http://lordoftheping.blogspot.com
>
>
> "Condor" <(E-Mail Removed)> wrote in message
> news:E52412A0-6A05-4371-A077-(E-Mail Removed)...
> > I've configured NLB on two Win 2k3 servers. These servers have two NIC's.
> > NLB
> > is configured on the front end lan (EFL) only. From a standalone server
> > which
> > has 'visibility' to this cluster - I'm able to ping both the individual
> > hosts, ping the EFL Cluster IP (by cluster name and IP). I ran a netmon
> > trace
> > and it indicated that my packets are being answered to. However, when I
> > try
> > to connect to this cluster (by name or IP) from the NLB manager of the
> > standalone server (I mentioned above) I get the message that "NLB is not
> > loacted on the specified computer". I have an application that needs to
> > connect via this "application NLB cluster" to login to the DB - which does
> > not go through. I'm beginning to believe that this has something to do
> > with
> > the above issue.
> >
> > Any pointers on how to solve this would be greatly appreciated.
> >
> > Thank you.
>
>

I forgot to add, my app's logs show the below error. The app is on the 3rd
server and it needs to connect to the below mentioned 'cluster' to access the
DB. I figured out that the connection to DB is not even coming through, since
it is not even making it to the cluster...

-2147023174 - The RPC server is unavailable. Looks like a system error but a
reboot did not help. I need to figure out what this means.

Thanks.

"Condor" wrote:

> Thanks Mat. I checked and the ports which my app needs are 80, 443 and 135 -
> these are already added to NLB config. NLB with 127.0.0.1 works good...
>
> I have an entry in the hosts file as well. 'UATRTCAPP' is the name of the
> cluster.
>
> 10.253.152.73 uatrtcapp
>
> I'm not a pro at this stuff but when I meant two network cards - I meant
> each node is on 2 LANS - front end (EFL) and back end (EBL). Clustering is
> done only on the EFL. Not sure how relevant it is to the current issue,
> though.
>
> Another thing is I can ping each node (from 3rd server) in the cluster and
> 'UATRTCAPP' from the third server which has visibility to the cluster. But
> when I try to connect to 'UATRTCAPP' from NLB manager I always get 'Could not
> locate NLB on specified computer'.
>
> One more interesting thing is when I reboot the third 'server' and
> immediately try to log into my app, it goes through. This is because NLB
> manager from the third server identifies 'UATRTCAPP' But after some time,
> login fails and NLB manager now gives 'Could not locate NLB.......' - same
> error as above. Looks like it is intermittent. I'm currently digging thru the
> Event viewer to see if I can find any errors.
>
> Hope my explanation is not confusing. Any pointers are greatly appreciated.
>
> "Mathieu CHATEAU" wrote:
>
> > Hello,
> >
> > I think you have two issues
> > -Access the NLB from the MMC
> > -access to a specific port on the NLB IP.
> >
> > As the NLB IP is binded and reachable, i think you forgot to add the needed
> > TCP port on the NLB, like the 1433 for SQL.
> >
> > On one node, click to add existing nlb. Type 127.0.0.1 as server to join.
> > Does it work ?
> >
> > Why are you speaking of two network card ?
> > Does the DB bind on all interfaces (netstat -an will help) ?
> >
> >
> > --
> > Cordialement,
> > Mathieu CHATEAU
> > http://lordoftheping.blogspot.com
> >
> >
> > "Condor" <(E-Mail Removed)> wrote in message
> > news:E52412A0-6A05-4371-A077-(E-Mail Removed)...
> > > I've configured NLB on two Win 2k3 servers. These servers have two NIC's.
> > > NLB
> > > is configured on the front end lan (EFL) only. From a standalone server
> > > which
> > > has 'visibility' to this cluster - I'm able to ping both the individual
> > > hosts, ping the EFL Cluster IP (by cluster name and IP). I ran a netmon
> > > trace
> > > and it indicated that my packets are being answered to. However, when I
> > > try
> > > to connect to this cluster (by name or IP) from the NLB manager of the
> > > standalone server (I mentioned above) I get the message that "NLB is not
> > > loacted on the specified computer". I have an application that needs to
> > > connect via this "application NLB cluster" to login to the DB - which does
> > > not go through. I'm beginning to believe that this has something to do
> > > with
> > > the above issue.
> > >
> > > Any pointers on how to solve this would be greatly appreciated.
> > >
> > > Thank you.
> >
> >

Are these 3 servers (2 NLB and the third) in a domain ?
Services are working through the nlb, it's just the nlb manager that you
want to get from remote ?

--
Cordialement,
Mathieu CHATEAU
http://lordoftheping.blogspot.com


"Condor" <(E-Mail Removed)> wrote in message
news:BA222BF2-56D3-463E-85F6-(E-Mail Removed)...
>I forgot to add, my app's logs show the below error. The app is on the 3rd
> server and it needs to connect to the below mentioned 'cluster' to access
> the
> DB. I figured out that the connection to DB is not even coming through,
> since
> it is not even making it to the cluster...
>
> -2147023174 - The RPC server is unavailable. Looks like a system error but
> a
> reboot did not help. I need to figure out what this means.
>
> Thanks.
>
> "Condor" wrote:
>
>> Thanks Mat. I checked and the ports which my app needs are 80, 443 and
>> 135 -
>> these are already added to NLB config. NLB with 127.0.0.1 works good...
>>
>> I have an entry in the hosts file as well. 'UATRTCAPP' is the name of
>> the
>> cluster.
>>
>> 10.253.152.73 uatrtcapp
>>
>> I'm not a pro at this stuff but when I meant two network cards - I meant
>> each node is on 2 LANS - front end (EFL) and back end (EBL). Clustering
>> is
>> done only on the EFL. Not sure how relevant it is to the current issue,
>> though.
>>
>> Another thing is I can ping each node (from 3rd server) in the cluster
>> and
>> 'UATRTCAPP' from the third server which has visibility to the cluster.
>> But
>> when I try to connect to 'UATRTCAPP' from NLB manager I always get 'Could
>> not
>> locate NLB on specified computer'.
>>
>> One more interesting thing is when I reboot the third 'server' and
>> immediately try to log into my app, it goes through. This is because NLB
>> manager from the third server identifies 'UATRTCAPP' But after some time,
>> login fails and NLB manager now gives 'Could not locate NLB.......' -
>> same
>> error as above. Looks like it is intermittent. I'm currently digging thru
>> the
>> Event viewer to see if I can find any errors.
>>
>> Hope my explanation is not confusing. Any pointers are greatly
>> appreciated.
>>
>> "Mathieu CHATEAU" wrote:
>>
>> > Hello,
>> >
>> > I think you have two issues
>> > -Access the NLB from the MMC
>> > -access to a specific port on the NLB IP.
>> >
>> > As the NLB IP is binded and reachable, i think you forgot to add the
>> > needed
>> > TCP port on the NLB, like the 1433 for SQL.
>> >
>> > On one node, click to add existing nlb. Type 127.0.0.1 as server to
>> > join.
>> > Does it work ?
>> >
>> > Why are you speaking of two network card ?
>> > Does the DB bind on all interfaces (netstat -an will help) ?
>> >
>> >
>> > --
>> > Cordialement,
>> > Mathieu CHATEAU
>> > http://lordoftheping.blogspot.com
>> >
>> >
>> > "Condor" <(E-Mail Removed)> wrote in message
>> > news:E52412A0-6A05-4371-A077-(E-Mail Removed)...
>> > > I've configured NLB on two Win 2k3 servers. These servers have two
>> > > NIC's.
>> > > NLB
>> > > is configured on the front end lan (EFL) only. From a standalone
>> > > server
>> > > which
>> > > has 'visibility' to this cluster - I'm able to ping both the
>> > > individual
>> > > hosts, ping the EFL Cluster IP (by cluster name and IP). I ran a
>> > > netmon
>> > > trace
>> > > and it indicated that my packets are being answered to. However, when
>> > > I
>> > > try
>> > > to connect to this cluster (by name or IP) from the NLB manager of
>> > > the
>> > > standalone server (I mentioned above) I get the message that "NLB is
>> > > not
>> > > loacted on the specified computer". I have an application that needs
>> > > to
>> > > connect via this "application NLB cluster" to login to the DB - which
>> > > does
>> > > not go through. I'm beginning to believe that this has something to
>> > > do
>> > > with
>> > > the above issue.
>> > >
>> > > Any pointers on how to solve this would be greatly appreciated.
>> > >
>> > > Thank you.
>> >
>> >

Yes, they are all in one domain. I just need to connect to 'UATRTCAPP'
cluster from the NLB of the third machine. If this goes through, the login
for my app also goes through. I am able to ping UATRTCAPP and it resolved to
the correct 10.253.x.y IP, though. I get the same 'Cannot locate NLB......'
error when I try to 'Connect to existing' from NLB manager.

Thanks.

"Mathieu CHATEAU" wrote:

> Are these 3 servers (2 NLB and the third) in a domain ?
> Services are working through the nlb, it's just the nlb manager that you
> want to get from remote ?
>
> --
> Cordialement,
> Mathieu CHATEAU
> http://lordoftheping.blogspot.com
>
>
> "Condor" <(E-Mail Removed)> wrote in message
> news:BA222BF2-56D3-463E-85F6-(E-Mail Removed)...
> >I forgot to add, my app's logs show the below error. The app is on the 3rd
> > server and it needs to connect to the below mentioned 'cluster' to access
> > the
> > DB. I figured out that the connection to DB is not even coming through,
> > since
> > it is not even making it to the cluster...
> >
> > -2147023174 - The RPC server is unavailable. Looks like a system error but
> > a
> > reboot did not help. I need to figure out what this means.
> >
> > Thanks.
> >
> > "Condor" wrote:
> >
> >> Thanks Mat. I checked and the ports which my app needs are 80, 443 and
> >> 135 -
> >> these are already added to NLB config. NLB with 127.0.0.1 works good...
> >>
> >> I have an entry in the hosts file as well. 'UATRTCAPP' is the name of
> >> the
> >> cluster.
> >>
> >> 10.253.152.73 uatrtcapp
> >>
> >> I'm not a pro at this stuff but when I meant two network cards - I meant
> >> each node is on 2 LANS - front end (EFL) and back end (EBL). Clustering
> >> is
> >> done only on the EFL. Not sure how relevant it is to the current issue,
> >> though.
> >>
> >> Another thing is I can ping each node (from 3rd server) in the cluster
> >> and
> >> 'UATRTCAPP' from the third server which has visibility to the cluster.
> >> But
> >> when I try to connect to 'UATRTCAPP' from NLB manager I always get 'Could
> >> not
> >> locate NLB on specified computer'.
> >>
> >> One more interesting thing is when I reboot the third 'server' and
> >> immediately try to log into my app, it goes through. This is because NLB
> >> manager from the third server identifies 'UATRTCAPP' But after some time,
> >> login fails and NLB manager now gives 'Could not locate NLB.......' -
> >> same
> >> error as above. Looks like it is intermittent. I'm currently digging thru
> >> the
> >> Event viewer to see if I can find any errors.
> >>
> >> Hope my explanation is not confusing. Any pointers are greatly
> >> appreciated.
> >>
> >> "Mathieu CHATEAU" wrote:
> >>
> >> > Hello,
> >> >
> >> > I think you have two issues
> >> > -Access the NLB from the MMC
> >> > -access to a specific port on the NLB IP.
> >> >
> >> > As the NLB IP is binded and reachable, i think you forgot to add the
> >> > needed
> >> > TCP port on the NLB, like the 1433 for SQL.
> >> >
> >> > On one node, click to add existing nlb. Type 127.0.0.1 as server to
> >> > join.
> >> > Does it work ?
> >> >
> >> > Why are you speaking of two network card ?
> >> > Does the DB bind on all interfaces (netstat -an will help) ?
> >> >
> >> >
> >> > --
> >> > Cordialement,
> >> > Mathieu CHATEAU
> >> > http://lordoftheping.blogspot.com
> >> >
> >> >
> >> > "Condor" <(E-Mail Removed)> wrote in message
> >> > news:E52412A0-6A05-4371-A077-(E-Mail Removed)...
> >> > > I've configured NLB on two Win 2k3 servers. These servers have two
> >> > > NIC's.
> >> > > NLB
> >> > > is configured on the front end lan (EFL) only. From a standalone
> >> > > server
> >> > > which
> >> > > has 'visibility' to this cluster - I'm able to ping both the
> >> > > individual
> >> > > hosts, ping the EFL Cluster IP (by cluster name and IP). I ran a
> >> > > netmon
> >> > > trace
> >> > > and it indicated that my packets are being answered to. However, when
> >> > > I
> >> > > try
> >> > > to connect to this cluster (by name or IP) from the NLB manager of
> >> > > the
> >> > > standalone server (I mentioned above) I get the message that "NLB is
> >> > > not
> >> > > loacted on the specified computer". I have an application that needs
> >> > > to
> >> > > connect via this "application NLB cluster" to login to the DB - which
> >> > > does
> >> > > not go through. I'm beginning to believe that this has something to
> >> > > do
> >> > > with
> >> > > the above issue.
> >> > >
> >> > > Any pointers on how to solve this would be greatly appreciated.
> >> > >
> >> > > Thank you.
> >> >
> >> >
>
>

did you try to connect to the nlb by using the real ip of one of the two
servers?

--
Cordialement,
Mathieu CHATEAU
http://lordoftheping.blogspot.com


"Condor" <(E-Mail Removed)> wrote in message
news:EDA252B4-2804-47CB-8C45-(E-Mail Removed)...
> Yes, they are all in one domain. I just need to connect to 'UATRTCAPP'
> cluster from the NLB of the third machine. If this goes through, the login
> for my app also goes through. I am able to ping UATRTCAPP and it resolved
> to
> the correct 10.253.x.y IP, though. I get the same 'Cannot locate
> NLB......'
> error when I try to 'Connect to existing' from NLB manager.
>
> Thanks.
>
> "Mathieu CHATEAU" wrote:
>
>> Are these 3 servers (2 NLB and the third) in a domain ?
>> Services are working through the nlb, it's just the nlb manager that you
>> want to get from remote ?
>>
>> --
>> Cordialement,
>> Mathieu CHATEAU
>> http://lordoftheping.blogspot.com
>>
>>
>> "Condor" <(E-Mail Removed)> wrote in message
>> news:BA222BF2-56D3-463E-85F6-(E-Mail Removed)...
>> >I forgot to add, my app's logs show the below error. The app is on the
>> >3rd
>> > server and it needs to connect to the below mentioned 'cluster' to
>> > access
>> > the
>> > DB. I figured out that the connection to DB is not even coming through,
>> > since
>> > it is not even making it to the cluster...
>> >
>> > -2147023174 - The RPC server is unavailable. Looks like a system error
>> > but
>> > a
>> > reboot did not help. I need to figure out what this means.
>> >
>> > Thanks.
>> >
>> > "Condor" wrote:
>> >
>> >> Thanks Mat. I checked and the ports which my app needs are 80, 443 and
>> >> 135 -
>> >> these are already added to NLB config. NLB with 127.0.0.1 works
>> >> good...
>> >>
>> >> I have an entry in the hosts file as well. 'UATRTCAPP' is the name of
>> >> the
>> >> cluster.
>> >>
>> >> 10.253.152.73 uatrtcapp
>> >>
>> >> I'm not a pro at this stuff but when I meant two network cards - I
>> >> meant
>> >> each node is on 2 LANS - front end (EFL) and back end (EBL).
>> >> Clustering
>> >> is
>> >> done only on the EFL. Not sure how relevant it is to the current
>> >> issue,
>> >> though.
>> >>
>> >> Another thing is I can ping each node (from 3rd server) in the cluster
>> >> and
>> >> 'UATRTCAPP' from the third server which has visibility to the cluster.
>> >> But
>> >> when I try to connect to 'UATRTCAPP' from NLB manager I always get
>> >> 'Could
>> >> not
>> >> locate NLB on specified computer'.
>> >>
>> >> One more interesting thing is when I reboot the third 'server' and
>> >> immediately try to log into my app, it goes through. This is because
>> >> NLB
>> >> manager from the third server identifies 'UATRTCAPP' But after some
>> >> time,
>> >> login fails and NLB manager now gives 'Could not locate NLB.......' -
>> >> same
>> >> error as above. Looks like it is intermittent. I'm currently digging
>> >> thru
>> >> the
>> >> Event viewer to see if I can find any errors.
>> >>
>> >> Hope my explanation is not confusing. Any pointers are greatly
>> >> appreciated.
>> >>
>> >> "Mathieu CHATEAU" wrote:
>> >>
>> >> > Hello,
>> >> >
>> >> > I think you have two issues
>> >> > -Access the NLB from the MMC
>> >> > -access to a specific port on the NLB IP.
>> >> >
>> >> > As the NLB IP is binded and reachable, i think you forgot to add the
>> >> > needed
>> >> > TCP port on the NLB, like the 1433 for SQL.
>> >> >
>> >> > On one node, click to add existing nlb. Type 127.0.0.1 as server to
>> >> > join.
>> >> > Does it work ?
>> >> >
>> >> > Why are you speaking of two network card ?
>> >> > Does the DB bind on all interfaces (netstat -an will help) ?
>> >> >
>> >> >
>> >> > --
>> >> > Cordialement,
>> >> > Mathieu CHATEAU
>> >> > http://lordoftheping.blogspot.com
>> >> >
>> >> >
>> >> > "Condor" <(E-Mail Removed)> wrote in message
>> >> > news:E52412A0-6A05-4371-A077-(E-Mail Removed)...
>> >> > > I've configured NLB on two Win 2k3 servers. These servers have two
>> >> > > NIC's.
>> >> > > NLB
>> >> > > is configured on the front end lan (EFL) only. From a standalone
>> >> > > server
>> >> > > which
>> >> > > has 'visibility' to this cluster - I'm able to ping both the
>> >> > > individual
>> >> > > hosts, ping the EFL Cluster IP (by cluster name and IP). I ran a
>> >> > > netmon
>> >> > > trace
>> >> > > and it indicated that my packets are being answered to. However,
>> >> > > when
>> >> > > I
>> >> > > try
>> >> > > to connect to this cluster (by name or IP) from the NLB manager of
>> >> > > the
>> >> > > standalone server (I mentioned above) I get the message that "NLB
>> >> > > is
>> >> > > not
>> >> > > loacted on the specified computer". I have an application that
>> >> > > needs
>> >> > > to
>> >> > > connect via this "application NLB cluster" to login to the DB -
>> >> > > which
>> >> > > does
>> >> > > not go through. I'm beginning to believe that this has something
>> >> > > to
>> >> > > do
>> >> > > with
>> >> > > the above issue.
>> >> > >
>> >> > > Any pointers on how to solve this would be greatly appreciated.
>> >> > >
>> >> > > Thank you.
>> >> >
>> >> >
>>
>>

NLB with real IP of the first server in the cluster goes through fine. If I
try to connect with the real IP of the second server, it throws "Could not
locate NLB on the specified computer". NLB is installed and configured on the
second server as well.

"Mathieu CHATEAU" wrote:

> did you try to connect to the nlb by using the real ip of one of the two
> servers?
>
> --
> Cordialement,
> Mathieu CHATEAU
> http://lordoftheping.blogspot.com
>
>
> "Condor" <(E-Mail Removed)> wrote in message
> news:EDA252B4-2804-47CB-8C45-(E-Mail Removed)...
> > Yes, they are all in one domain. I just need to connect to 'UATRTCAPP'
> > cluster from the NLB of the third machine. If this goes through, the login
> > for my app also goes through. I am able to ping UATRTCAPP and it resolved
> > to
> > the correct 10.253.x.y IP, though. I get the same 'Cannot locate
> > NLB......'
> > error when I try to 'Connect to existing' from NLB manager.
> >
> > Thanks.
> >
> > "Mathieu CHATEAU" wrote:
> >
> >> Are these 3 servers (2 NLB and the third) in a domain ?
> >> Services are working through the nlb, it's just the nlb manager that you
> >> want to get from remote ?
> >>
> >> --
> >> Cordialement,
> >> Mathieu CHATEAU
> >> http://lordoftheping.blogspot.com
> >>
> >>
> >> "Condor" <(E-Mail Removed)> wrote in message
> >> news:BA222BF2-56D3-463E-85F6-(E-Mail Removed)...
> >> >I forgot to add, my app's logs show the below error. The app is on the
> >> >3rd
> >> > server and it needs to connect to the below mentioned 'cluster' to
> >> > access
> >> > the
> >> > DB. I figured out that the connection to DB is not even coming through,
> >> > since
> >> > it is not even making it to the cluster...
> >> >
> >> > -2147023174 - The RPC server is unavailable. Looks like a system error
> >> > but
> >> > a
> >> > reboot did not help. I need to figure out what this means.
> >> >
> >> > Thanks.
> >> >
> >> > "Condor" wrote:
> >> >
> >> >> Thanks Mat. I checked and the ports which my app needs are 80, 443 and
> >> >> 135 -
> >> >> these are already added to NLB config. NLB with 127.0.0.1 works
> >> >> good...
> >> >>
> >> >> I have an entry in the hosts file as well. 'UATRTCAPP' is the name of
> >> >> the
> >> >> cluster.
> >> >>
> >> >> 10.253.152.73 uatrtcapp
> >> >>
> >> >> I'm not a pro at this stuff but when I meant two network cards - I
> >> >> meant
> >> >> each node is on 2 LANS - front end (EFL) and back end (EBL).
> >> >> Clustering
> >> >> is
> >> >> done only on the EFL. Not sure how relevant it is to the current
> >> >> issue,
> >> >> though.
> >> >>
> >> >> Another thing is I can ping each node (from 3rd server) in the cluster
> >> >> and
> >> >> 'UATRTCAPP' from the third server which has visibility to the cluster.
> >> >> But
> >> >> when I try to connect to 'UATRTCAPP' from NLB manager I always get
> >> >> 'Could
> >> >> not
> >> >> locate NLB on specified computer'.
> >> >>
> >> >> One more interesting thing is when I reboot the third 'server' and
> >> >> immediately try to log into my app, it goes through. This is because
> >> >> NLB
> >> >> manager from the third server identifies 'UATRTCAPP' But after some
> >> >> time,
> >> >> login fails and NLB manager now gives 'Could not locate NLB.......' -
> >> >> same
> >> >> error as above. Looks like it is intermittent. I'm currently digging
> >> >> thru
> >> >> the
> >> >> Event viewer to see if I can find any errors.
> >> >>
> >> >> Hope my explanation is not confusing. Any pointers are greatly
> >> >> appreciated.
> >> >>
> >> >> "Mathieu CHATEAU" wrote:
> >> >>
> >> >> > Hello,
> >> >> >
> >> >> > I think you have two issues
> >> >> > -Access the NLB from the MMC
> >> >> > -access to a specific port on the NLB IP.
> >> >> >
> >> >> > As the NLB IP is binded and reachable, i think you forgot to add the
> >> >> > needed
> >> >> > TCP port on the NLB, like the 1433 for SQL.
> >> >> >
> >> >> > On one node, click to add existing nlb. Type 127.0.0.1 as server to
> >> >> > join.
> >> >> > Does it work ?
> >> >> >
> >> >> > Why are you speaking of two network card ?
> >> >> > Does the DB bind on all interfaces (netstat -an will help) ?
> >> >> >
> >> >> >
> >> >> > --
> >> >> > Cordialement,
> >> >> > Mathieu CHATEAU
> >> >> > http://lordoftheping.blogspot.com
> >> >> >
> >> >> >
> >> >> > "Condor" <(E-Mail Removed)> wrote in message
> >> >> > news:E52412A0-6A05-4371-A077-(E-Mail Removed)...
> >> >> > > I've configured NLB on two Win 2k3 servers. These servers have two
> >> >> > > NIC's.
> >> >> > > NLB
> >> >> > > is configured on the front end lan (EFL) only. From a standalone
> >> >> > > server
> >> >> > > which
> >> >> > > has 'visibility' to this cluster - I'm able to ping both the
> >> >> > > individual
> >> >> > > hosts, ping the EFL Cluster IP (by cluster name and IP). I ran a
> >> >> > > netmon
> >> >> > > trace
> >> >> > > and it indicated that my packets are being answered to. However,
> >> >> > > when
> >> >> > > I
> >> >> > > try
> >> >> > > to connect to this cluster (by name or IP) from the NLB manager of
> >> >> > > the
> >> >> > > standalone server (I mentioned above) I get the message that "NLB
> >> >> > > is
> >> >> > > not
> >> >> > > loacted on the specified computer". I have an application that
> >> >> > > needs
> >> >> > > to
> >> >> > > connect via this "application NLB cluster" to login to the DB -
> >> >> > > which
> >> >> > > does
> >> >> > > not go through. I'm beginning to believe that this has something
> >> >> > > to
> >> >> > > do
> >> >> > > with
> >> >> > > the above issue.
> >> >> > >
> >> >> > > Any pointers on how to solve this would be greatly appreciated.
> >> >> > >
> >> >> > > Thank you.
> >> >> >
> >> >> >
> >>
> >>
>
>

does this server has netbios removed or disabled ?
Are you in unicast with one network card ?


--
Cordialement,
Mathieu CHATEAU
http://lordoftheping.blogspot.com


"Condor" <(E-Mail Removed)> wrote in message
news:6765D93E-3735-4DE2-9D69-(E-Mail Removed)...
> NLB with real IP of the first server in the cluster goes through fine. If
> I
> try to connect with the real IP of the second server, it throws "Could not
> locate NLB on the specified computer". NLB is installed and configured on
> the
> second server as well.
>
> "Mathieu CHATEAU" wrote:
>
>> did you try to connect to the nlb by using the real ip of one of the two
>> servers?
>>
>> --
>> Cordialement,
>> Mathieu CHATEAU
>> http://lordoftheping.blogspot.com
>>
>>
>> "Condor" <(E-Mail Removed)> wrote in message
>> news:EDA252B4-2804-47CB-8C45-(E-Mail Removed)...
>> > Yes, they are all in one domain. I just need to connect to 'UATRTCAPP'
>> > cluster from the NLB of the third machine. If this goes through, the
>> > login
>> > for my app also goes through. I am able to ping UATRTCAPP and it
>> > resolved
>> > to
>> > the correct 10.253.x.y IP, though. I get the same 'Cannot locate
>> > NLB......'
>> > error when I try to 'Connect to existing' from NLB manager.
>> >
>> > Thanks.
>> >
>> > "Mathieu CHATEAU" wrote:
>> >
>> >> Are these 3 servers (2 NLB and the third) in a domain ?
>> >> Services are working through the nlb, it's just the nlb manager that
>> >> you
>> >> want to get from remote ?
>> >>
>> >> --
>> >> Cordialement,
>> >> Mathieu CHATEAU
>> >> http://lordoftheping.blogspot.com
>> >>
>> >>
>> >> "Condor" <(E-Mail Removed)> wrote in message
>> >> news:BA222BF2-56D3-463E-85F6-(E-Mail Removed)...
>> >> >I forgot to add, my app's logs show the below error. The app is on
>> >> >the
>> >> >3rd
>> >> > server and it needs to connect to the below mentioned 'cluster' to
>> >> > access
>> >> > the
>> >> > DB. I figured out that the connection to DB is not even coming
>> >> > through,
>> >> > since
>> >> > it is not even making it to the cluster...
>> >> >
>> >> > -2147023174 - The RPC server is unavailable. Looks like a system
>> >> > error
>> >> > but
>> >> > a
>> >> > reboot did not help. I need to figure out what this means.
>> >> >
>> >> > Thanks.
>> >> >
>> >> > "Condor" wrote:
>> >> >
>> >> >> Thanks Mat. I checked and the ports which my app needs are 80, 443
>> >> >> and
>> >> >> 135 -
>> >> >> these are already added to NLB config. NLB with 127.0.0.1 works
>> >> >> good...
>> >> >>
>> >> >> I have an entry in the hosts file as well. 'UATRTCAPP' is the name
>> >> >> of
>> >> >> the
>> >> >> cluster.
>> >> >>
>> >> >> 10.253.152.73 uatrtcapp
>> >> >>
>> >> >> I'm not a pro at this stuff but when I meant two network cards - I
>> >> >> meant
>> >> >> each node is on 2 LANS - front end (EFL) and back end (EBL).
>> >> >> Clustering
>> >> >> is
>> >> >> done only on the EFL. Not sure how relevant it is to the current
>> >> >> issue,
>> >> >> though.
>> >> >>
>> >> >> Another thing is I can ping each node (from 3rd server) in the
>> >> >> cluster
>> >> >> and
>> >> >> 'UATRTCAPP' from the third server which has visibility to the
>> >> >> cluster.
>> >> >> But
>> >> >> when I try to connect to 'UATRTCAPP' from NLB manager I always get
>> >> >> 'Could
>> >> >> not
>> >> >> locate NLB on specified computer'.
>> >> >>
>> >> >> One more interesting thing is when I reboot the third 'server' and
>> >> >> immediately try to log into my app, it goes through. This is
>> >> >> because
>> >> >> NLB
>> >> >> manager from the third server identifies 'UATRTCAPP' But after some
>> >> >> time,
>> >> >> login fails and NLB manager now gives 'Could not locate
>> >> >> NLB.......' -
>> >> >> same
>> >> >> error as above. Looks like it is intermittent. I'm currently
>> >> >> digging
>> >> >> thru
>> >> >> the
>> >> >> Event viewer to see if I can find any errors.
>> >> >>
>> >> >> Hope my explanation is not confusing. Any pointers are greatly
>> >> >> appreciated.
>> >> >>
>> >> >> "Mathieu CHATEAU" wrote:
>> >> >>
>> >> >> > Hello,
>> >> >> >
>> >> >> > I think you have two issues
>> >> >> > -Access the NLB from the MMC
>> >> >> > -access to a specific port on the NLB IP.
>> >> >> >
>> >> >> > As the NLB IP is binded and reachable, i think you forgot to add
>> >> >> > the
>> >> >> > needed
>> >> >> > TCP port on the NLB, like the 1433 for SQL.
>> >> >> >
>> >> >> > On one node, click to add existing nlb. Type 127.0.0.1 as server
>> >> >> > to
>> >> >> > join.
>> >> >> > Does it work ?
>> >> >> >
>> >> >> > Why are you speaking of two network card ?
>> >> >> > Does the DB bind on all interfaces (netstat -an will help) ?
>> >> >> >
>> >> >> >
>> >> >> > --
>> >> >> > Cordialement,
>> >> >> > Mathieu CHATEAU
>> >> >> > http://lordoftheping.blogspot.com
>> >> >> >
>> >> >> >
>> >> >> > "Condor" <(E-Mail Removed)> wrote in message
>> >> >> > news:E52412A0-6A05-4371-A077-(E-Mail Removed)...
>> >> >> > > I've configured NLB on two Win 2k3 servers. These servers have
>> >> >> > > two
>> >> >> > > NIC's.
>> >> >> > > NLB
>> >> >> > > is configured on the front end lan (EFL) only. From a
>> >> >> > > standalone
>> >> >> > > server
>> >> >> > > which
>> >> >> > > has 'visibility' to this cluster - I'm able to ping both the
>> >> >> > > individual
>> >> >> > > hosts, ping the EFL Cluster IP (by cluster name and IP). I ran
>> >> >> > > a
>> >> >> > > netmon
>> >> >> > > trace
>> >> >> > > and it indicated that my packets are being answered to.
>> >> >> > > However,
>> >> >> > > when
>> >> >> > > I
>> >> >> > > try
>> >> >> > > to connect to this cluster (by name or IP) from the NLB manager
>> >> >> > > of
>> >> >> > > the
>> >> >> > > standalone server (I mentioned above) I get the message that
>> >> >> > > "NLB
>> >> >> > > is
>> >> >> > > not
>> >> >> > > loacted on the specified computer". I have an application that
>> >> >> > > needs
>> >> >> > > to
>> >> >> > > connect via this "application NLB cluster" to login to the DB -
>> >> >> > > which
>> >> >> > > does
>> >> >> > > not go through. I'm beginning to believe that this has
>> >> >> > > something
>> >> >> > > to
>> >> >> > > do
>> >> >> > > with
>> >> >> > > the above issue.
>> >> >> > >
>> >> >> > > Any pointers on how to solve this would be greatly appreciated.
>> >> >> > >
>> >> >> > > Thank you.
>> >> >> >
>> >> >> >
>> >>
>> >>
>>
>>

Yes, NETBIOS is disabled (per security policy on the EFL which is internet
facing) but is enabled on the EBL - on both the servers. Each server has 2
NIC's (one for EFL and one for EBL) and clustering is in unicast mode on both
servers.

"Mathieu CHATEAU" wrote:

> does this server has netbios removed or disabled ?
> Are you in unicast with one network card ?
>
>
> --
> Cordialement,
> Mathieu CHATEAU
> http://lordoftheping.blogspot.com
>
>
> "Condor" <(E-Mail Removed)> wrote in message
> news:6765D93E-3735-4DE2-9D69-(E-Mail Removed)...
> > NLB with real IP of the first server in the cluster goes through fine. If
> > I
> > try to connect with the real IP of the second server, it throws "Could not
> > locate NLB on the specified computer". NLB is installed and configured on
> > the
> > second server as well.
> >
> > "Mathieu CHATEAU" wrote:
> >
> >> did you try to connect to the nlb by using the real ip of one of the two
> >> servers?
> >>
> >> --
> >> Cordialement,
> >> Mathieu CHATEAU
> >> http://lordoftheping.blogspot.com
> >>
> >>
> >> "Condor" <(E-Mail Removed)> wrote in message
> >> news:EDA252B4-2804-47CB-8C45-(E-Mail Removed)...
> >> > Yes, they are all in one domain. I just need to connect to 'UATRTCAPP'
> >> > cluster from the NLB of the third machine. If this goes through, the
> >> > login
> >> > for my app also goes through. I am able to ping UATRTCAPP and it
> >> > resolved
> >> > to
> >> > the correct 10.253.x.y IP, though. I get the same 'Cannot locate
> >> > NLB......'
> >> > error when I try to 'Connect to existing' from NLB manager.
> >> >
> >> > Thanks.
> >> >
> >> > "Mathieu CHATEAU" wrote:
> >> >
> >> >> Are these 3 servers (2 NLB and the third) in a domain ?
> >> >> Services are working through the nlb, it's just the nlb manager that
> >> >> you
> >> >> want to get from remote ?
> >> >>
> >> >> --
> >> >> Cordialement,
> >> >> Mathieu CHATEAU
> >> >> http://lordoftheping.blogspot.com
> >> >>
> >> >>
> >> >> "Condor" <(E-Mail Removed)> wrote in message
> >> >> news:BA222BF2-56D3-463E-85F6-(E-Mail Removed)...
> >> >> >I forgot to add, my app's logs show the below error. The app is on
> >> >> >the
> >> >> >3rd
> >> >> > server and it needs to connect to the below mentioned 'cluster' to
> >> >> > access
> >> >> > the
> >> >> > DB. I figured out that the connection to DB is not even coming
> >> >> > through,
> >> >> > since
> >> >> > it is not even making it to the cluster...
> >> >> >
> >> >> > -2147023174 - The RPC server is unavailable. Looks like a system
> >> >> > error
> >> >> > but
> >> >> > a
> >> >> > reboot did not help. I need to figure out what this means.
> >> >> >
> >> >> > Thanks.
> >> >> >
> >> >> > "Condor" wrote:
> >> >> >
> >> >> >> Thanks Mat. I checked and the ports which my app needs are 80, 443
> >> >> >> and
> >> >> >> 135 -
> >> >> >> these are already added to NLB config. NLB with 127.0.0.1 works
> >> >> >> good...
> >> >> >>
> >> >> >> I have an entry in the hosts file as well. 'UATRTCAPP' is the name
> >> >> >> of
> >> >> >> the
> >> >> >> cluster.
> >> >> >>
> >> >> >> 10.253.152.73 uatrtcapp
> >> >> >>
> >> >> >> I'm not a pro at this stuff but when I meant two network cards - I
> >> >> >> meant
> >> >> >> each node is on 2 LANS - front end (EFL) and back end (EBL).
> >> >> >> Clustering
> >> >> >> is
> >> >> >> done only on the EFL. Not sure how relevant it is to the current
> >> >> >> issue,
> >> >> >> though.
> >> >> >>
> >> >> >> Another thing is I can ping each node (from 3rd server) in the
> >> >> >> cluster
> >> >> >> and
> >> >> >> 'UATRTCAPP' from the third server which has visibility to the
> >> >> >> cluster.
> >> >> >> But
> >> >> >> when I try to connect to 'UATRTCAPP' from NLB manager I always get
> >> >> >> 'Could
> >> >> >> not
> >> >> >> locate NLB on specified computer'.
> >> >> >>
> >> >> >> One more interesting thing is when I reboot the third 'server' and
> >> >> >> immediately try to log into my app, it goes through. This is
> >> >> >> because
> >> >> >> NLB
> >> >> >> manager from the third server identifies 'UATRTCAPP' But after some
> >> >> >> time,
> >> >> >> login fails and NLB manager now gives 'Could not locate
> >> >> >> NLB.......' -
> >> >> >> same
> >> >> >> error as above. Looks like it is intermittent. I'm currently
> >> >> >> digging
> >> >> >> thru
> >> >> >> the
> >> >> >> Event viewer to see if I can find any errors.
> >> >> >>
> >> >> >> Hope my explanation is not confusing. Any pointers are greatly
> >> >> >> appreciated.
> >> >> >>
> >> >> >> "Mathieu CHATEAU" wrote:
> >> >> >>
> >> >> >> > Hello,
> >> >> >> >
> >> >> >> > I think you have two issues
> >> >> >> > -Access the NLB from the MMC
> >> >> >> > -access to a specific port on the NLB IP.
> >> >> >> >
> >> >> >> > As the NLB IP is binded and reachable, i think you forgot to add
> >> >> >> > the
> >> >> >> > needed
> >> >> >> > TCP port on the NLB, like the 1433 for SQL.
> >> >> >> >
> >> >> >> > On one node, click to add existing nlb. Type 127.0.0.1 as server
> >> >> >> > to
> >> >> >> > join.
> >> >> >> > Does it work ?
> >> >> >> >
> >> >> >> > Why are you speaking of two network card ?
> >> >> >> > Does the DB bind on all interfaces (netstat -an will help) ?
> >> >> >> >
> >> >> >> >
> >> >> >> > --
> >> >> >> > Cordialement,
> >> >> >> > Mathieu CHATEAU
> >> >> >> > http://lordoftheping.blogspot.com
> >> >> >> >
> >> >> >> >
> >> >> >> > "Condor" <(E-Mail Removed)> wrote in message
> >> >> >> > news:E52412A0-6A05-4371-A077-(E-Mail Removed)...
> >> >> >> > > I've configured NLB on two Win 2k3 servers. These servers have
> >> >> >> > > two
> >> >> >> > > NIC's.
> >> >> >> > > NLB
> >> >> >> > > is configured on the front end lan (EFL) only. From a
> >> >> >> > > standalone
> >> >> >> > > server
> >> >> >> > > which
> >> >> >> > > has 'visibility' to this cluster - I'm able to ping both the
> >> >> >> > > individual
> >> >> >> > > hosts, ping the EFL Cluster IP (by cluster name and IP). I ran
> >> >> >> > > a
> >> >> >> > > netmon
> >> >> >> > > trace
> >> >> >> > > and it indicated that my packets are being answered to.
> >> >> >> > > However,
> >> >> >> > > when
> >> >> >> > > I
> >> >> >> > > try
> >> >> >> > > to connect to this cluster (by name or IP) from the NLB manager
> >> >> >> > > of
> >> >> >> > > the
> >> >> >> > > standalone server (I mentioned above) I get the message that
> >> >> >> > > "NLB
> >> >> >> > > is
> >> >> >> > > not
> >> >> >> > > loacted on the specified computer". I have an application that
> >> >> >> > > needs
> >> >> >> > > to
> >> >> >> > > connect via this "application NLB cluster" to login to the DB -
> >> >> >> > > which
> >> >> >> > > does
> >> >> >> > > not go through. I'm beginning to believe that this has
> >> >> >> > > something
> >> >> >> > > to
> >> >> >> > > do
> >> >> >> > > with
> >> >> >> > > the above issue.
> >> >> >> > >
> >> >> >> > > Any pointers on how to solve this would be greatly appreciated.
> >> >> >> > >
> >> >> >> > > Thank you.
> >> >> >> >
> >> >> >> >
> >> >>
> >> >>
> >>
> >>
>
>