Showing posts with label scenario. Show all posts
Showing posts with label scenario. Show all posts

Wednesday, March 28, 2012

model db during restore

Hello, i have no specific problem but i just want to know your opinion on the issue. Last week I tested my recovery scenario and after installing new server i successfully restored master db. Then I tried to start server, but it wouldn't come up because physical path of model db in master..sysdatabases table didn't correspond to the actual location of db on new server, so that tempdb couldn't be created from model db on server startup. I started server without recovering dbs with exception of master using trace -T3608. Now I hoped that I could restore model using MOVE clause but it gave me error saying that restore operation needs to use temporary table in tempdb. To break this vicious circle I used model db files from installation CD and put them to the expected location. After two restarts server accepted the files and I went on restoring msdb and user databases. But I'm sure I must be missing something and there must be better way to do this but i haven't been able to find anything relating to this problem. Thank you for your input. mojzaGo to Microsoft site and download the best practices anylyzer. Its very good for variables read on the database and could help in some other ways.

Wednesday, March 7, 2012

Missing replicated records

Hi gurus there,
I have a puzzling scenario here.
Environment: SQL Server 2K5 (sp2) EE + windows 2K3
I have a publisher server PubSrv_1and two subscriber severs SubSrv_1 and
SubSrv_2, there are two publications on the Publication_1, and
Publication_2, but they all contain the same table as article tbl_A
Publication_1 is a transational replication with Subscription on SubSrv_1
and Publication_2 is a transactional replication with updateable
subscription on SubSrv_2.
Issue: sometimes I find there are missing records which appear in tbl_A on
PubSrv_1 but not in tbl_A on SubSrv_2 while still in tbl_A on SubSrv_1. In
another word, some records in the publication table are NOT replicated to
one subscriber but still replicated to another subscriber.
I am confused how this happens. Could someone please shed some lights here?
TIA
Jeff
Can you check the conflicts log on the subscriber to see if the
missing rows are logged there?
On Jan 9, 3:03 am, "jeff_y" <jeffrey...@.abc.com> wrote:
> Hi gurus there,
> I have a puzzling scenario here.
> Environment: SQL Server 2K5 (sp2) EE + windows 2K3
> I have a publisher server PubSrv_1and two subscriber severs SubSrv_1 and
> SubSrv_2, there are two publications on the Publication_1, and
> Publication_2, but they all contain the same table as article tbl_A
> Publication_1 is a transational replication with Subscription on SubSrv_1
> and Publication_2 is a transactional replication with updateable
> subscription on SubSrv_2.
> Issue: sometimes I find there are missing records which appear in tbl_A on
> PubSrv_1 but not in tbl_A on SubSrv_2 while still in tbl_A on SubSrv_1. In
> another word, some records in the publication table are NOT replicated to
> one subscriber but still replicated to another subscriber.
> I am confused how this happens. Could someone please shed some lights here?
> TIA
> Jeff
|||Hi Hillary,
Thanks for your quick response.
I am not sure where the conflicts log are for the transactional replication.
More information:
I have even put a delete trigger on the problematic table at the subsriber
side to try to log any deletion on the table, but find the trigger has never
been fired off. So in another word, the missing records have never been
deleted as they were in fact never replicated over from the publisher side.
I find the two subscribers have different editions, the SubSrv_2, which has
seen the issues, has edition 3042, while SubSrv_1 has 3161. Do you think
this can cause the issue (I am not convinced though)?
Thanks,
Jeff
"Hilary Cotter" <hilary.cotter@.gmail.com> wrote in message
news:86aa0601-0caa-49ad-b9a1-d5d3cc732456@.f10g2000hsf.googlegroups.com...
> Can you check the conflicts log on the subscriber to see if the
> missing rows are logged there?
> On Jan 9, 3:03 am, "jeff_y" <jeffrey...@.abc.com> wrote:
>