Bitten in butt by DFS / FRS problem!

  • Thread starter Thread starter Daniel Billingsley
  • Start date Start date
D

Daniel Billingsley

I've had two dfs replicas (with frs replication set up and verified working)
for a particular folder running for quite some time.

Last Friday morning one of the target servers had a memory problem and
rebooted several times. I eventually took that target offline so users
would not connect to it.

But several users reported that particular files had resorted to versions
that were a few weeks old.

As best I've been able to determine, ServerA -> ServerB replication was
still working, but ServerB -> ServerA replication had stopped some time ago,
presumably a few weeks ago based on the symptoms from the users. The
replication settings in the XP DFS manager still look fine, there are no
errors or warnings in the logs of either server.

So, I think the scenario was that for the users that received the ServerB
reference and were saving their files there, those changes were not
replicated back to ServerA, so when those users got "transferred" to the
ServerA reference they were looking at old versions of files.

My question is just simply if anyone else has experienced this. I don't
really necessarily expect to get to the bottom of it.

One of things I need to decide at this point is if DFS/FRS can really be
relied upon to provide the high availability I need or if I need to start
looking at 3rd party solutions.
 
Hi,
I will follow this up with the FRS team.
Thanks

--
Thanks,
Arif Saifee [MSFT]

PS: Please post DFS related queries in the newsgroup
"microsoft.public.win2000.file_system". Please use "DFS" in the subject to
make it immediately noticeable.

Disclaimer: This posting is provided "AS IS" with no warranties, and confers
no rights.
 
Not that I really expect anything, but did you ever hear anything regarding
this Arif... other than the "he's crazy, that can't happen" that I expect?
:)

Arif Saifee said:
Hi,
I will follow this up with the FRS team.
Thanks

--
Thanks,
Arif Saifee [MSFT]

PS: Please post DFS related queries in the newsgroup
"microsoft.public.win2000.file_system". Please use "DFS" in the subject to
make it immediately noticeable.

Disclaimer: This posting is provided "AS IS" with no warranties, and confers
no rights.

Daniel Billingsley said:
I've had two dfs replicas (with frs replication set up and verified working)
for a particular folder running for quite some time.

Last Friday morning one of the target servers had a memory problem and
rebooted several times. I eventually took that target offline so users
would not connect to it.

But several users reported that particular files had resorted to versions
that were a few weeks old.

As best I've been able to determine, ServerA -> ServerB replication was
still working, but ServerB -> ServerA replication had stopped some time ago,
presumably a few weeks ago based on the symptoms from the users. The
replication settings in the XP DFS manager still look fine, there are no
errors or warnings in the logs of either server.

So, I think the scenario was that for the users that received the ServerB
reference and were saving their files there, those changes were not
replicated back to ServerA, so when those users got "transferred" to the
ServerA reference they were looking at old versions of files.

My question is just simply if anyone else has experienced this. I don't
really necessarily expect to get to the bottom of it.

One of things I need to decide at this point is if DFS/FRS can really be
relied upon to provide the high availability I need or if I need to start
looking at 3rd party solutions.
Click to expand...
Click to expand...
 
Daniel,
As you theorize, it does sound like ServerB -> ServerA replication could be
broken. A quick test would be to drop a small text file into ServerB & see
if it makes it to ServerA (and vice versa). Anyway, at this point I would
ask:

1) Are there any errors or warnings in the eventlog for FRS?
2) Are you using Sonar to monitor FRS? If so, is it reporting any errors,
replication backlogs, etc?
3) How widespread is the issue? Is it specific to only a few files or
widespread among all replicated content? This would indicate different
problems of course..

Anyway, let us know. If you haven't seen the FRS troubleshooting package w/
Sonar, you can get it here:

http://www.microsoft.com/windows2000/techinfo/reskit/tools/new/sonar-o.asp

--
Brian [MSFT]

Disclaimer: This posting is provided "AS IS" with no warranties, and confers
no rights. Please do not send e-mail directly to this alias. This alias is
for newsgroup purposes only.

Daniel Billingsley said:
Not that I really expect anything, but did you ever hear anything regarding
this Arif... other than the "he's crazy, that can't happen" that I expect?
:)

Arif Saifee said:
Hi,
I will follow this up with the FRS team.
Thanks

--
Thanks,
Arif Saifee [MSFT]

PS: Please post DFS related queries in the newsgroup
"microsoft.public.win2000.file_system". Please use "DFS" in the subject to
make it immediately noticeable.

Disclaimer: This posting is provided "AS IS" with no warranties, and confers
no rights.

I've had two dfs replicas (with frs replication set up and verified working)
for a particular folder running for quite some time.

Last Friday morning one of the target servers had a memory problem and
rebooted several times. I eventually took that target offline so users
would not connect to it.

But several users reported that particular files had resorted to versions
that were a few weeks old.

As best I've been able to determine, ServerA -> ServerB replication was
still working, but ServerB -> ServerA replication had stopped some
Click to expand...
time
ago,
presumably a few weeks ago based on the symptoms from the users. The
replication settings in the XP DFS manager still look fine, there are no
errors or warnings in the logs of either server.

So, I think the scenario was that for the users that received the ServerB
reference and were saving their files there, those changes were not
replicated back to ServerA, so when those users got "transferred" to the
ServerA reference they were looking at old versions of files.

My question is just simply if anyone else has experienced this. I don't
really necessarily expect to get to the bottom of it.

One of things I need to decide at this point is if DFS/FRS can really be
relied upon to provide the high availability I need or if I need to start
looking at 3rd party solutions.
Click to expand...
Click to expand...
Click to expand...
 
Hi Daniel,

I have redirected your query to the FRS team. Someone from the team should
be replying to you shortly
--
Thanks,
Arif Saifee [MSFT]

PS: Please post DFS related queries in the newsgroup
"microsoft.public.win2000.file_system". Please use "DFS" in the subject to
make it immediately noticeable.

Disclaimer: This posting is provided "AS IS" with no warranties, and confers
no rights.

Daniel Billingsley said:
Not that I really expect anything, but did you ever hear anything regarding
this Arif... other than the "he's crazy, that can't happen" that I expect?
:)

Arif Saifee said:
Hi,
I will follow this up with the FRS team.
Thanks

--
Thanks,
Arif Saifee [MSFT]

PS: Please post DFS related queries in the newsgroup
"microsoft.public.win2000.file_system". Please use "DFS" in the subject to
make it immediately noticeable.

Disclaimer: This posting is provided "AS IS" with no warranties, and confers
no rights.

I've had two dfs replicas (with frs replication set up and verified working)
for a particular folder running for quite some time.

Last Friday morning one of the target servers had a memory problem and
rebooted several times. I eventually took that target offline so users
would not connect to it.

But several users reported that particular files had resorted to versions
that were a few weeks old.

As best I've been able to determine, ServerA -> ServerB replication was
still working, but ServerB -> ServerA replication had stopped some
Click to expand...
time
ago,
presumably a few weeks ago based on the symptoms from the users. The
replication settings in the XP DFS manager still look fine, there are no
errors or warnings in the logs of either server.

So, I think the scenario was that for the users that received the ServerB
reference and were saving their files there, those changes were not
replicated back to ServerA, so when those users got "transferred" to the
ServerA reference they were looking at old versions of files.

My question is just simply if anyone else has experienced this. I don't
really necessarily expect to get to the bottom of it.

One of things I need to decide at this point is if DFS/FRS can really be
relied upon to provide the high availability I need or if I need to start
looking at 3rd party solutions.
Click to expand...
Click to expand...
Click to expand...
 
That's what I did to verify it Brian. I put File1 on ServerA and it quickly
appears on ServerB. Put File2 on ServerB and... nothing. Change File1 on
ServerA and it's quickly changed on ServerB. Change File1 on ServerB and
ServerA is never modified. No messages in the event logs whatsoever.

Unfortunately for any future testing or debugging, I've ripped this replica
out of my DFS configuration, partly in a rage and partly because I need to
move on. My main question was really if anyone else had experienced this.
I was relying on DFS/FRS to provide reliable high availability but it turned
out it in fact to be worse than no redundancy at all in this case. Needless
to say, I am VERY hesitant to rely on this technology now.

Brian said:
Daniel,
As you theorize, it does sound like ServerB -> ServerA replication could be
broken. A quick test would be to drop a small text file into ServerB & see
if it makes it to ServerA (and vice versa). Anyway, at this point I would
ask:

1) Are there any errors or warnings in the eventlog for FRS?
2) Are you using Sonar to monitor FRS? If so, is it reporting any errors,
replication backlogs, etc?
3) How widespread is the issue? Is it specific to only a few files or
widespread among all replicated content? This would indicate different
problems of course..

Anyway, let us know. If you haven't seen the FRS troubleshooting package w/
Sonar, you can get it here:

http://www.microsoft.com/windows2000/techinfo/reskit/tools/new/sonar-o.asp

--
Brian [MSFT]

Disclaimer: This posting is provided "AS IS" with no warranties, and confers
no rights. Please do not send e-mail directly to this alias. This alias is
for newsgroup purposes only.

Daniel Billingsley said:
Not that I really expect anything, but did you ever hear anything regarding
this Arif... other than the "he's crazy, that can't happen" that I expect?
:)

Arif Saifee said:
Hi,
I will follow this up with the FRS team.
Thanks

--
Thanks,
Arif Saifee [MSFT]

PS: Please post DFS related queries in the newsgroup
"microsoft.public.win2000.file_system". Please use "DFS" in the
Click to expand...
Click to expand...
subject
to
Click to expand...
are
no
Click to expand...
really
be
Click to expand...
Click to expand...
 
Daniel,

It is hard to say why your DFS structure's FRS replication was not
working without any troubleshooting data, logs or further information. There
are scenarios where if FRS is not configured correctly or if one of its
dependencies are broken (ie: network connectivity, name resolution, AD
replication) then FRS may not work. If you'd like to further look into it
feel free to post back and we'll help as much as possible.

Regards,
~Luiz (MSFT)


--
Please do not send e-mail directly to this alias. This alias is for
newsgroup purposes only.


Daniel Billingsley said:
That's what I did to verify it Brian. I put File1 on ServerA and it quickly
appears on ServerB. Put File2 on ServerB and... nothing. Change File1 on
ServerA and it's quickly changed on ServerB. Change File1 on ServerB and
ServerA is never modified. No messages in the event logs whatsoever.

Unfortunately for any future testing or debugging, I've ripped this replica
out of my DFS configuration, partly in a rage and partly because I need to
move on. My main question was really if anyone else had experienced this.
I was relying on DFS/FRS to provide reliable high availability but it turned
out it in fact to be worse than no redundancy at all in this case. Needless
to say, I am VERY hesitant to rely on this technology now.

Brian said:
Daniel,
As you theorize, it does sound like ServerB -> ServerA replication could be
broken. A quick test would be to drop a small text file into ServerB & see
if it makes it to ServerA (and vice versa). Anyway, at this point I would
ask:

1) Are there any errors or warnings in the eventlog for FRS?
2) Are you using Sonar to monitor FRS? If so, is it reporting any errors,
replication backlogs, etc?
3) How widespread is the issue? Is it specific to only a few files or
widespread among all replicated content? This would indicate different
problems of course..

Anyway, let us know. If you haven't seen the FRS troubleshooting
Click to expand...
package
w/
Sonar, you can get it here:

http://www.microsoft.com/windows2000/techinfo/reskit/tools/new/sonar-o.asp

--
Brian [MSFT]

Disclaimer: This posting is provided "AS IS" with no warranties, and confers
no rights. Please do not send e-mail directly to this alias. This alias is
for newsgroup purposes only.

Not that I really expect anything, but did you ever hear anything regarding
this Arif... other than the "he's crazy, that can't happen" that I expect?
:)

Hi,
I will follow this up with the FRS team.
Thanks

--
Thanks,
Arif Saifee [MSFT]

PS: Please post DFS related queries in the newsgroup
"microsoft.public.win2000.file_system". Please use "DFS" in the
Click to expand...
subject
to
make it immediately noticeable.

Disclaimer: This posting is provided "AS IS" with no warranties, and
confers
no rights.

I've had two dfs replicas (with frs replication set up and verified
working)
for a particular folder running for quite some time.

Last Friday morning one of the target servers had a memory problem and
rebooted several times. I eventually took that target offline so users
would not connect to it.

But several users reported that particular files had resorted to
versions
that were a few weeks old.

As best I've been able to determine, ServerA -> ServerB
Click to expand...
replication
was
still working, but ServerB -> ServerA replication had stopped some time
ago,
presumably a few weeks ago based on the symptoms from the users. The
replication settings in the XP DFS manager still look fine, there
Click to expand...
are
no
errors or warnings in the logs of either server.

So, I think the scenario was that for the users that received the
ServerB
reference and were saving their files there, those changes were not
replicated back to ServerA, so when those users got "transferred"
Click to expand...
to
the
ServerA reference they were looking at old versions of files.

My question is just simply if anyone else has experienced this. I don't
really necessarily expect to get to the bottom of it.

One of things I need to decide at this point is if DFS/FRS can
Click to expand...
really
be
relied upon to provide the high availability I need or if I need to
start
looking at 3rd party solutions.
Click to expand...
Click to expand...
Click to expand...
 
Back
Top