| Groups | Blog | Home |
sql server replication : Initial SnapShot failure....
I have a series of NT4 systems that I need to set up replication on.
They have sp 6a installed on them I have done a clean install of SQL 2k Server then added sp3a Setting up the publication provides no problems, but when I try to run the snapshot agent, I have been recieving problmes (different on different systems :) and all have exactly the same hardware/software. I should also mention that I have it running well on 2 of 5. The issues seem to be either that I have an "incompatable stub"... or just a generic "suspect agent" error. Neither error points to a specific file or stored proc. Any ideas welcome!!!! Thanks Rob
Hi Rob,
What you described sounds like some sort of DCOM configuration problem. Either the replication agent COM servers are not registered properly on some of your machines or DCOM is disabled\improperly configured on them. To diagnose the issue further, you may want to try the following: 1) Try running snapshot.exe manually from the command line; snapshot.exe can be found at %ProgramFiles% \Microsoft SQL Server\80\COM. If running snapshot.exe without arguments displays the usage screen, chances are good that it is a DCOM activation issue. To further confirm that this is indeed the case, you can try running the other replication agents such as distrib.exe, logread.exe, qrdrsvc.exe, and replmerg.exe from the same location. 2) Try starting the other replication agents from enterprise manager and see if they fail with the same symptoms as the snapshot agent. If so, chances are real good that you have a general DCOM activation problem. 3) If you think that you have a general DCOM activation problem, check out the system\application event log and see if you can find any messages that will point you in the right direction of what the problem may be. 4) As a way to give you a better feel of what the problem may be, you should open taskmanager when you try to start any replication agent from enterprise manager and see if the corresponding replication agent process (snapshot.exe,distrib.exe, ...etc.) is started. If you suspect that the replication agents are not registered properly, you can manually registered them by doing the following from %ProgramFiles%\Microsoft SQL Server\80 (you may want to try it anyway): regsvr32 replagnt.dll snapshot.exe /RegServer distrib.exe /RegServer replmerg.exe /RegServer logread.exe /RegServer qrdrsvc.exe /RegServer If nothing that I have suggested so far works for you, you should check out the Microsoft KB for some of the more arcane DCOM activation issues. HTH -Raymond [quoted text, click to view] >-----Original Message-----
>I have a series of NT4 systems that I need to set up replication on. >They have sp 6a installed on them >I have done a clean install of SQL 2k Server then added sp3a > >Setting up the publication provides no problems, but when I try to run the >snapshot agent, I have been recieving problmes (different on different >systems :) and all have exactly the same hardware/software. > >I should also mention that I have it running well on 2 of 5. > >The issues seem to be either that I have an "incompatable stub"... or just a >generic "suspect agent" error. > >Neither error points to a specific file or stored proc. > >Any ideas welcome!!!! > >Thanks Rob > >.
Raymond, Thanks, I will try what you suggest.. (it was hitting me as dcom or
a dll issue, but I hate to do a file compare on all the dlls :) Hillary, no error numbers :) it would be nice, but the errors are too generic and have no additional information. I will post back and let you know how it pans out.. Rob PS After running an @@version on 4 of the effected servers the I got the results below. (SVR005 is the only one that is working :( ? ) No difference in the install process. Running command on svr005 ; Microsoft SQL Server 2000 - 8.00.760 (Intel X86) Dec 17 2002 14:22:05 Copyright (c) 1988-2003 Microsoft Corporation Standard Edition on Windows NT 4.0 (Build 1381: Service Pack 6) Running command on svr001 ; Microsoft SQL Server 2000 - 8.00.760 (Intel X86) Dec 17 2002 14:22:05 Copyright (c) 1988-2003 Microsoft Corporation Standard Edition on Windows NT 4.0 (Build 1381: Service Pack 6) Running command on svr010 ; Microsoft SQL Server 2000 - 8.00.760 (Intel X86) Dec 17 2002 14:22:05 Copyright (c) 1988-2003 Microsoft Corporation Standard Edition on Windows NT 4.0 (Build 1381: Service Pack 6) Running command on svr083 ; Microsoft SQL Server 2000 - 8.00.760 (Intel X86) Dec 17 2002 14:22:05 Copyright (c) 1988-2003 Microsoft Corporation Standard Edition on Windows NT 4.0 (Build 1381: Service Pack 6) [quoted text, click to view] "Raymond Mak [MSFT]" wrote:
> Hi Rob, > > What you described sounds like some sort of DCOM > configuration problem. Either the replication agent COM > servers are not registered properly on some of your > machines or DCOM is disabled\improperly configured on > them. To diagnose the issue further, you may want to try > the following: > > 1) Try running snapshot.exe manually from the command > line; snapshot.exe can be found at %ProgramFiles% > \Microsoft SQL Server\80\COM. If running snapshot.exe > without arguments displays the usage screen, chances are > good that it is a DCOM activation issue. To further > confirm that this is indeed the case, you can try running > the other replication agents such as distrib.exe, > logread.exe, qrdrsvc.exe, and replmerg.exe from the same > location. > 2) Try starting the other replication agents from > enterprise manager and see if they fail with the same > symptoms as the snapshot agent. If so, chances are real > good that you have a general DCOM activation problem. > 3) If you think that you have a general DCOM activation > problem, check out the system\application event log and > see if you can find any messages that will point you in > the right direction of what the problem may be. > 4) As a way to give you a better feel of what the problem > may be, you should open taskmanager when you try to start > any replication agent from enterprise manager and see if > the corresponding replication agent process > (snapshot.exe,distrib.exe, ...etc.) is started. > > If you suspect that the replication agents are not > registered properly, you can manually registered them by > doing the following from %ProgramFiles%\Microsoft SQL > Server\80 (you may want to try it anyway): > > regsvr32 replagnt.dll > snapshot.exe /RegServer > distrib.exe /RegServer > replmerg.exe /RegServer > logread.exe /RegServer > qrdrsvc.exe /RegServer > > If nothing that I have suggested so far works for you, you > should check out the Microsoft KB for some of the more > arcane DCOM activation issues. > > HTH > > -Raymond > > >-----Original Message----- > >I have a series of NT4 systems that I need to set up > replication on. > >They have sp 6a installed on them > >I have done a clean install of SQL 2k Server then added > sp3a > > > >Setting up the publication provides no problems, but when > I try to run the > >snapshot agent, I have been recieving problmes (different > on different > >systems :) and all have exactly the same > hardware/software. > > > >I should also mention that I have it running well on 2 of > 5. > > > >The issues seem to be either that I have an "incompatable > stub"... or just a > >generic "suspect agent" error. > > > >Neither error points to a specific file or stored proc. > > > >Any ideas welcome!!!! > > > >Thanks Rob > > > >. > >
Ok, maybe getting further....
(I registered all exe's and dll's in then COM directory - dug into domcnfg to no real value - really didn't know what to look for there but teh settings seemed to be the same as where things are working) Now I am not seeing the RPC stub error.. What seems to happen now is that I can start the Snapshot agent, but it timeouts(after an interval of 4 min) I have just cranked the failure time up to 15 min to see if this helps.. I will post back on success of failure. Any other ideas welcome (I really don't want to do the replication manually, but I need to institute a replicating subscriber model on 156 servers from 21 other servers in the next week and at this point using sql's replication engine is being a ^*^*&^* pain.)
Ok things are getting more interesting....
When I run the Snapshot from the command prompt I get an instant snapshot (Using the systemadmin sql login - but running from em gives me a stub error - after a good long timeout) I am about to go home and bring down the bumps with some good cold beer.. Let you know how things go tomorrow Rob MSCE (nt4) MSCD v6 and net MCDBA 7 & 2k
can you give me the error message and error number associated with the
incompatible stub message? the suspect agent problem can be fixed by right clicking on Replication Monitor and selecting Refresh Rate and Settings, Then set Inactivity Threshold to something larger. You MUST monitor your agents to ensure that they are doing something on the subscriber - as they could be hung. -- Hilary Cotter Looking for a SQL Server replication book? http://www.nwsu.com/0974973602.html Looking for a FAQ on Indexing Services/SQL FTS http://www.indexserverfaq.com [quoted text, click to view] "rob lynch" <roblynch@discussions.microsoft.com> wrote in message news:44436678-59E3-4B74-A390-AE5F9E3369AC@microsoft.com... > I have a series of NT4 systems that I need to set up replication on. > They have sp 6a installed on them > I have done a clean install of SQL 2k Server then added sp3a > > Setting up the publication provides no problems, but when I try to run the > snapshot agent, I have been recieving problmes (different on different > systems :) and all have exactly the same hardware/software. > > I should also mention that I have it running well on 2 of 5. > > The issues seem to be either that I have an "incompatable stub"... or just a > generic "suspect agent" error. > > Neither error points to a specific file or stored proc. > > Any ideas welcome!!!! > > Thanks Rob >
Getting Closer (no beer yet!)
EM fails but Command line always works.... Suggestions? TIA
Rob, did you see anything interesting in the NT event log?
Did the snapshot agent ever pop up in task manager when you start it from EM anyway? One cheap workaround that you can use is to run the replication agent from the CMD subsystem of SQLServerAgent but that is strictly the last resort. HTH [quoted text, click to view] >-----Original Message-----
>Getting Closer (no beer yet!) > >EM fails but Command line always works.... > >Suggestions? > >TIA > >Rob >.
Rob, there are a few more things that you can try:
1) If you haven't done this after you re-register the dlls, stop and restart SQLServerAgent. 2) What is the service account of SQLServerAgent for the machines that are giving you trouble? 3) Do you have any kind of fire wall running that may affect DCOM\RPC traffic? [quoted text, click to view] >-----Original Message-----
>Getting Closer (no beer yet!) > >EM fails but Command line always works.... > >Suggestions? > >TIA > >Rob >.
Ok... (sorry went home had a beer and did some sleeping.. Didn't help the
servers, but I feel better :) Raymond, I have tried stopping and restarting the sqlserveragent and that had no effect, the event log shows nothing special - just the incompatible stub message (message at the bottom). I don't think it could be a firewall issue as the problem shows up when you log on locally (but there is no firewall there anyway.) The account used by SQLServer the SQLServerAgent are both a member of the domain/administrators group. I am wondering if it could be an NT4 with SQL2K issue, but both are running the latest sp's, and of course 1 is working fine :) Do you know where or how sql runs its replication tasks. Is it via a stored proc call or does it run via some internal dll code? By the way, thanks for all your help so far. It is nice to have a fresh set of brains working on this issue.
Forgot the message :)
SubSystem Message - Job 'SVR083-Northwind-083LN01E-1' (0xE090DB0AB794D911AC8700E0B8151568), step 2 - Incompatible version of the
Ok so now it is getting ugly :)
I compared the files in two of the servers bin and the COM directorys and have found some small differences.. Wondering if this could be the issue.. There are also some files on the one that doesn't work that are missing on the one that does, but I doubt that will have much of an impact... At this moment I have tired moving cmdwrap.exe over and also the contents of the com dir (and reregistered + stop and start service) still no go... Rob (server that works) (Binn) 04/17/01 11:22p 21,056 cmdwrap.exe 04/27/00 07:25p 4,516 sqlctr.h 04/27/00 07:25p 19,984 sqlctr.ini 08/05/00 08:51p 32,825 sqlboot.dll 12/17/02 05:24p 25,152 opends60.dll 12/17/02 05:25p 33,336 ssmsad70.dll 12/17/02 05:25p 33,336 SSmsVI70.dll 12/17/02 05:25p 33,344 SSmsSH70.dll 12/17/02 05:26p 25,152 xpsqlbot.dll (Com) 08/06/00 11:17a 91,648 replerrx.dll 08/06/00 01:51a 98,376 repldts.dll 08/06/00 01:50a 28,745 repldsui.dll (server that doesn't) (BINN) 04/28/00 02:25a 4,516 sqlctr.h 04/28/00 02:25a 19,984 sqlctr.ini 04/28/00 12:06a 4,576 schema.txt 08/06/00 01:50a 20,545 cmdwrap.exe 08/06/00 01:50a 20,555 ftsetup.exe 08/06/00 01:50a 24,639 opends60.dll 08/06/00 01:51a 24,643 xpsqlbot.dll 08/06/00 01:51a 32,823 ssmsad70.dll 08/06/00 01:51a 32,823 ssmsvi70.dll 08/06/00 01:51a 32,825 sqlboot.dll 08/06/00 01:51a 32,830 ssmssh70.dll 12/17/02 05:25p 197,196 sqlftqry.dll (COM) 12/17/02 05:24p 33,352 repldsui.dll 12/17/02 05:24p 98,888 repldts.dll
Well that probably falls under the category of nice try :(
I went throught all the sql direcotorys and looked for file date differences and copyied them over then registered what I could then tried again.. Same result! Latest thinking... If the builtin command fails (replciation command) but the command shell verison works, then I will change all the replication tasks to use xp_cmdshell.. YUK... wil post back results....
Rob, a search for the error message on the microsoft web
site yields some very interesting results: http://search.microsoft.com/search/results.aspx? st=b&na=88&View=en- us&qu=Incompatible+version+of+the+RPC+stub Most of the links returned mention something along the line of mismatched OLE\COM dlls installed by various programs. I would strongly encourage you to check those out and see if they help. [quoted text, click to view] >-----Original Message-----
>Forgot the message :) > >SubSystem Message - Job 'SVR083-Northwind-083LN01E-1' >(0xE090DB0AB794D911AC8700E0B8151568), step 2 - Incompatible version of the >RPC stub >.
Final result / solution...
Chaseing down COM dll errors with nonspecific pointers/errors was a hugly frustrating waste of time.. (Never did find it, reinstalled sp3, upgraded to MDAC 2.8.... Bottom line, a bust!~ What worked (and seems quite reliable - but I wil continue to watch it) Step1. Setup Distributer Step2. Setup Publisher Setp3. Setup Publication Step4. (critical) - run exec sp_changepublication @publication='northwind', @property='independent_agent', @value='TRUE' (Where northwind is the name of your publication) Step5. Setup your subscribers - don't sync or initialize the publication --- Below here is where the automatic agent would fail ---- Step6. Create a task (or execute in QA) the following SQLStatement (replace your servername etc, all following examples use a publication of northwind and everything else about it is northwind...) exec master..xp_cmdshell '"c:\Program Files\Microsoft SQL Server\80\COM\SNAPSHOT.EXE" -Publisher yourPublicationServerName -PublisherDB Northwind -Publication northwind -Distributor yourDistributionServerName-DistributorSecurityMode 1 ' Note DistributorSecurityMode 1 sets up Integrated Secutity This will successfully create the initial Snapshot.. Step7. Publish your Snapshot (make sure the database exists first - on the subscriber, this processess won't create a db, just the tables and data) exec master..xp_cmdshell '"c:\Program Files\Microsoft SQL Server\80\COM\Distrib.EXE" -subscriberDB northwind -Subscriber YourSubscriber -Publisher YourPublisher -PublisherDB Northwind -Publication northwind -Distributor YourDistributer -DistributorSecurityMode 1 ' Now the data will exist in the remote location (you will need 1 statement per subscriber - a good thing to script) Step8. Run a log reader to make see if there is any new data to move... exec master..xp_cmdshell '"c:\Program Files\Microsoft SQL Server\80\COM\LogRead.EXE" -Publisher YourPublisher -PublisherDB Northwind -Distributor YourDistributer -DistributorSecurityMode 1 ' At that point you just setup tasks to alternate between step 8 and 7.. YOu can read the return of 8 and based on transactions existing execute 7) Phew!L@O&*^&^$ What a long 4 days!! Rob PS gratitude goes to fmpia@hotmail.com who has a great resource on replication http://www.geocities.com/priyadarshu/tech/Repl.htm also SQL Server 2000 Security - Part 8 - Introduction to Replication Security By Marcin Policht http://www.databasejournal.com/features/mssql/article.php/3383221 Brad Syputa - MS http://www.windowsforumz.com/-incompatible-rpc-stub-ftopict188785.html who started my thinking into xp_Cmdshell... And last, but certainly NOT LEAST Raymond Mak.. Thanks for pointing me at the Agents in the COM directory. This was the Breakthrought that enabled me to finally put it all together!!!!
OOps
1 last thing on reread.. step 4 should be run in the context of the database being replicated... ie use Nortwind go xp_cmdshell 'C:\Prog.......'
Raymond Mak...
YOU DA MAN!! Raymond, I got a good night sleep and then chased a lead you gave me yesterday. It pointed to a file called mcrepair.exe.. It came from one of the MSDN articles and strangly enough is a "Microsoft Money" tool. I ran it this morning and Bingo!! replce oleaut and a couple more and we were good to go. Of course now I neeed to work out what I broke when I replaced those udpated and later files :) But I don't care. Now I have replciation working!!!
Don't see what you're looking for? Try a search.
|
April 2003
June 2003
July 2003
August 2003
September 2003
October 2003
November 2003
December 2003
January 2004
February 2004
March 2004
April 2004
May 2004
June 2004
July 2004
August 2004
September 2004
October 2004
November 2004
December 2004
January 2005
February 2005
March 2005
April 2005
May 2005
June 2005
July 2005
August 2005
September 2005
October 2005
November 2005
December 2005
January 2006
February 2006
March 2006
April 2006
May 2006
June 2006
July 2006
August 2006
September 2006
October 2006
November 2006
December 2006
January 2007
February 2007
March 2007
April 2007
May 2007
June 2007
July 2007
August 2007
September 2007
October 2007
November 2007
October 2008
| home · blog · groups | Questions? Comments? Contact the dn |