OT: DFS/FRS Problem

Hi guys,

I've asked this question in the dfs_frs group but have had no reply as of
yet. I thought someone here might be able to help!

We have just setup a 2nd server ('Server2', win2k server sp4), which is
running DFS to provide fault tolerance to our main server. I setup about 20
or so DFS links, with DFS replicas. I.e. \\server1\sales which is master,
and \\server2\sales which was added as a replica. So far its replicated
about 15 or so of the links from server1 to server2 however its now just
stopped, and I can't get it to replicate anymore. The remaining 5 links,
have a few files inside which look to have been replicated, but not
everything. During the replication process I set SERVER1 as a root replica
of SERVER2, to make sure it was fully fault tolerant.

In the event log I'm getting: Event ID: 13508 Event Source: NtFrs Computer:
SERVER2.
There are 4 or 5 entries like this, 1 for each of the 5 folders it
hasn't replicated i.e. e:\data\rsgfiles e:\data\sales e:\data\it etc

Description:
The File Replication Service is having trouble enabling replication from
SERVER1 to SERVER2 for e:\data\rsgfiles using the DNS name
SERVER1.ourdomain.gb. FRS will keep retrying.
Following are some of the reasons you would see this warning.

[1] FRS can not correctly resolve the DNS name SERVER1.ourdomain.gb from
this computer.
[2] FRS is not running on SERVER1.ourdomain.gb.
[3] The topology information in the Active Directory for this replica has
not yet replicated to all the Domain Controllers.

This event log message will appear once per connection, After the problem
is fixed you will see another event log message indicating that the
connection has been established.

Now I can ping SERVER1 using its FQDN so I'm sure DNS is fine, FRS is
running on SERVER1, I've even stopped and restarted it, I'm not sure about
the topology information, it should have replicated, we only have 3 DCs (all
win2k) and all in the same room. I have looked on eventid.net for 13508 &
NtFrs and tried some of the idea there, including NTFRSUTL VERSION
SERVER1.ourdomain.gb to verify that FRS is running on it, I've run netdiag
/fix to correct missing DNS SRV records. And I've also re-installed SP4, on
it just to make sure.

Could setting SERVER1 as a root replica to SERVER2 have caused this? I was
getting a few errors saying the staging area was full which I think was due
to having some very large files (over 650mb). So I changed the regisry value
HKLM\SYSTEM\CurrentControlSet\Services\Ntfrs\Param eters\Staging Space Limit
in KB to double the decimal value. Could this have caused it?

Any help greatly appreciated.

--
Ben

IT Professional, MCP 70-210
"On my way to becoming fully certifiable!"

"The internet - where men are men, women are men and children are FBI
agents."

Ben

incase you missed it "Ben" <> said in
news:#:

> In the event log I'm getting: Event ID: 13508 Event Source: NtFrs
> Computer: SERVER2.
> There are 4 or 5 entries like this, 1 for each of the 5 folders it
> hasn't replicated i.e. e:\data\rsgfiles e:\data\sales e:\data\it etc

13508 is an RPC error being unable to make a secure connection to a
server in the replica mix. you might want to fire up RPCping and give it
a whirl. check to see if there is a NTFRSAPI.LOG file in teh %systemroot%
\debug folder. It might shed some light on it as well...

--
Neil MCNGP #30
the "curious" hair on the soap of society

Neil

incase you missed it "Ben" <> said in
news:#:

> Could setting SERVER1 as a root replica to SERVER2 have caused this? I
> was getting a few errors saying the staging area was full which I
> think was due to having some very large files (over 650mb). So I
> changed the regisry value
> HKLM\SYSTEM\CurrentControlSet\Services\Ntfrs\Param eters\Staging Space
> Limit in KB to double the decimal value. Could this have caused it?
>

if the staging directory is full you will get a 13522 error (staging
directory is full) or a 13511 error (database is out of disk space). the
reg hack was exaclty what you were suppose to do and doubling the amount
(assuming you do have sufficient space) should be ok

--
Neil MCNGP #30
the "curious" hair on the soap of society

Neil

"Neil" <guess!!!@gmail.com> wrote in message
news:Xns958284EA83ADneilmcsegmailcom@207.46.248.16 ...
> incase you missed it "Ben" <> said in
> news:#:
>
> > In the event log I'm getting: Event ID: 13508 Event Source: NtFrs
> > Computer: SERVER2.
> > There are 4 or 5 entries like this, 1 for each of the 5 folders it
> > hasn't replicated i.e. e:\data\rsgfiles e:\data\sales e:\data\it etc
>
> 13508 is an RPC error being unable to make a secure connection to a
> server in the replica mix. you might want to fire up RPCping and give it
> a whirl. check to see if there is a NTFRSAPI.LOG file in teh %systemroot%
> \debug folder. It might shed some light on it as well...
>
> --
> Neil MCNGP #30
> the "curious" hair on the soap of society

There are 5x NtFrs_000x.log (1-5) files which have a modified date of today,
looking at them doesn't really make much sense. The NtFrsApi.log that is
there has a modified date of 4 days ago, which was when I 1st set it up and
looking at it, there is nothing recent inside.
I have downloaded RPC Ping, and will try running it to see if it gives any
errors.
Cheers
Ben

Ben

"Neil" <guess!!!@gmail.com> wrote in message
news:Xns958284EA83ADneilmcsegmailcom@207.46.248.16 ...
>
> 13508 is an RPC error being unable to make a secure connection to a
> server in the replica mix. you might want to fire up RPCping and give it
> a whirl. check to see if there is a NTFRSAPI.LOG file in teh %systemroot%
> \debug folder. It might shed some light on it as well...
>
> --
> Neil MCNGP #30
> the "curious" hair on the soap of society

I have also just run dcdiag on both servers, which has give some errors.

From SERVER1

<snip>
Testing server: Default-First-Site-Name\GBSERV1-2K
Starting test: Replications
[Replications Check,GBSERV1-2K] A recent replication attempt
failed:
From GBSERV2-2K to GBSERV1-2K
Naming Context:
CN=Schema,CN=Configuration,DC=triumph-international,DC=gb
The replication generated an error (1722):
Win32 Error 1722
The failure occurred at 2004-10-15 08:46.38.
The last success occurred at 2004-10-15 07:45.35.
1 failures have occurred since the last success.
The source GBSERV2-2K is responding now.
......................... GBSERV1-2K passed test Replications
<snip>
Starting test: systemlog
An Error Event occured. EventID: 0x40011006
Time Generated: 10/15/2004 08:48:33
Event String: The connection was aborted by the remote WINS.

An Error Event occured. EventID: 0xC0011093
Time Generated: 10/15/2004 09:02:32
Event String: WINS Pull thread encountered an error during the

......................... GBSERV1-2K failed test systemlog
<snip>

From SERVER2

<snip>
Starting test: systemlog
An Error Event occured. EventID: 0x800009CF
Time Generated: 10/15/2004 08:48:11
Event String: The server service was unable to recreate the

An Error Event occured. EventID: 0x800009CF
Time Generated: 10/15/2004 08:48:11
Event String: The server service was unable to recreate the

......................... GBSERV2-2K failed test systemlog
<snip>

I have deleted the entire DFS root, and recreated it this morning with just
3 test folders. Its now reliplicating, however before doing so it gave the
same DNS errors. Will see what rpc ping says now!

Ben

Ben

incase you missed it "Ben" <> said in
news::

> Starting test: systemlog
> An Error Event occured. EventID: 0x800009CF
> Time Generated: 10/15/2004 08:48:11
> Event String: The server service was unable to recreate the
>
> An Error Event occured. EventID: 0x800009CF
> Time Generated: 10/15/2004 08:48:11
> Event String: The server service was unable to recreate the
>
> ......................... GBSERV2-2K failed test systemlog

might be worth having a look at kb839499 and 319504 to see if there is
anything that starts you thinking. it's what shows up when I googled
using the following search:
"failed test systemlog" site:microsoft.com

HTH

--
Neil MCNGP #30
the "curious" hair on the soap of society

Neil

In article <#>,
says...
> Now I can ping SERVER1 using its FQDN so I'm sure DNS is fine,
>
That does not mean that DNS is fine. You need to have all of the other
appropriate records besides the A record.

Just an FYI.

Laura

Laura A. Robinson