a quick guide to some of the more common commands you’ll need;
create connection to phoebe.sundry
set error class rs_sqlserver_error_class
set function string class rs_function_string_class
set username sundry_maint
set password sundry_maint_ps
with log transfer on
The log transfer on means that it will be able to replicate data as well as receive it.
Also rememebr that you have to add the database to the replication server by using rs_init in the install directory. This installs the stored procs and tables into the db as needed by the repserver.
Create replication definition
create replication definition rep1
with primary at phoebe.sundry
with all tables named ‘alpha1’ (a int, b char(10))
primary key (a)
create sybscription sub1
with replicate at phoebe.sundry2
check subscription sub1
With replicate at phoebe.sundry2
You can also use the command rs_helpsub from the RSSD of the repserver.
This restarts the dsi threads etc
resume connection to phoebe.sundry2
drop subscription sub1
with replicate at phoebe.sundry2
This sometimes doesn’t work if this is the case then go into the RSSD database for the replication server and delete the relevant line from the rs_subscriptions table, you have to do this if you want to drop the connection.
This will tell you what is up or down etc
Tells you specifically what isn’t working
If it’s a connection then try to suspend the connection and resume it, if it’s a repagent then use sp_stop_rep_agent <dbname> from the dataserver and then sp_start_rep_agent <dbname> to get i8t working again, if that fails check the error log.
This is a program which can be found in the bin directory of the Replication server installation, it is used to resyncronise the replicate database with the primary database, after a system failure, the sysadmin manual has an exstensive description of its use.
The best way to run this is to create a config file for it to run against. The config file will look something like;
PDS = TOKYO_DS
RDS = SYDNEY_DS
PDB = pubs2
RDB = pubs2
PTABLE = titleauthor
RTABLE = titleauthor
PSELECT = select au_id, title_id, au_ord, royaltyper from titleauthor order by au_id,\ title_id
RSELECT = select au_id, title_id, au_ord, royaltyper from titleauthor order by au_id,\ title_id
PUSER = repuser
RUSER = repuser
PPWD = piglet
RPWD = piglet
KEY = au_id
KEY = title_id
RECONCILE = Y
VISUAL = Y
NUM_TRIES = 3
WAIT = 10
If this file was called reconcile1.cfg then the command would be
rs_subcmp –f reconcile1.cfg
If you want to reconcile the primary database from the replicate database then you would swap around the PDB and RDB and also the PDS and RDS. You could either have a separate file for this or use command line options, -d for the replicate database and –D for the primary database and then –s for the replicate dataserver and –S for the primary dataserver.
This command will find the transaction which failed to be replicated, it’s run from the replication server controlling the database with the problem. It writes the output to the exceptions log. This consists of 3 tables in the RSSD for the Repserver, rs_excepthdr, rs_exceptcmd and rs_exceptslast. These tables should be purged periodically after having fixed the problem. You use rs_helpexception to see a list of the exceptions stored in the RSSD. And to delete it you would use rs_delexception.