Well, this question is to Karim 
I've just added that if you want to see the output of the dbcc command on your screen you need to run DBCC TRACEON(3604) first.
But I think it's because of
"The DBTABLE structure has an output parameter called dbt_open. This parameter keeps track of how many users are in the database"
/Annie
Annie,
To add your point, there is one more parameter dbt_keep from the dbcc output.
dbcc dbtable(dbname)
go
any value greater than 0 for dbt_keep indicates that there are some users connected to database or dbt_keep count did not get cleanup properly.
Regards
Kiran K Adharapuram
Hi Kiran,
What about other parameters ??
Just let me know.
Regards
Hi Marco,
both of your servers are in stanby mode. This is usually the case when you start the servers in the wrong sequence.
I assume that your failover succeeded and now you standby is actually the primary server.
We can try this out by restarting the system again, but now start XX1main first and then XX1standby.
If we end up in the same situation then we could force one of the servers to become primary. The only problem might be if that if there is still data in the SRS queue the systems will not be in sync afterwards. IN this case I would need all the logs ASE/SRS/DR_agent from both servers to figure out what happened during the failover.
Regards,
Joachim
We'll need a bit more detailed info on the commands you're issuing as well as your disk and network environment.
I did a write-up on how dumps/loads are processed several years ago. While that post is probably available on some archived/read-only copy of the old sybase newsgroup forums, I'm unable (at the moment) to locate said forums. The following link should suffice:
http://codeverge.com/sybase.ase.backup+recovery/database-load-from-a-compressed-du/967855
Check out the section on background info on how the backupserver (really the 'sybmultbuf' processes work). In a nutshell ...
- compressed dumps (and loads) require cpu cycles, so make sure you're not maxing out cpu cycles
- dumping across a network (including dumping to locally mounted NFS drives) is usually slower than dumping to a local disk drive
And in this thread ...Speeding up full backup of Replicate database ASE 15.5 ... you'll see that you also need to pay attention to the disk subsystem throughput on your machines ... throughput for database devices as well as the filesystem(s) where the dump devices reside.
Hi Joachim,
i tried restarting the server but i always got the same result, no primary_site. How can I force the server to become primary? The server is just for testing purposes, so we don't mind if there is a data loss.
Best Regards
Marco
Hello Mark,
thx for the link, I will go through.
Which information do you need exactly?
Maybe we are focusing more on CPU-item, as the FS I think isn't the bottleneck, as it is the same FS.
And if the network is the bottleneck, the performance-issue has to work the other way, thats the weird thing.
So I think we might focus more on CPU utilization.
Can I check the utilization of dump-process somewhere there?
Do I have to focus on "sybmultbul"?
So on the 1st view (via top) the sybmultbuf seems to be equal from utilization.
I also forgot.
I execute direct command "dump database <SID> to "<directory>"
If you're not performing compressed dumps then you can likely ignore the sybmultbuf processes since their cpu usage will be quite minimal (ie, without compression the sybmultbuf processes spend most of the time waiting for disk reads/writes to complete).
Ok, but what else might be the problem or what can I check?
Might it be an performance-leak on reading the data?
CPU utilization is likely a non-issue if you're not using compression.
re: the FS ...
Is the FS directly mounted to both machines in the same manner, and with the same configurations/settings? For example, is the FS mounted on one machine with journaling enabled, but journaling disabled on the other machine?
re: disks in general ...
I'd suggest having your systems folks monitor all disk operations (database device reads/writes, dump device/FS reads writes) to see if there are any bottlenecks.
re: network ...
Are your 2 machines by any chance on different networks, or perhaps in different computer centers? Even if you're mounting the same FS, with the same configurations/settings, a 'slow' network will (obviously) cause slowdowns.
-----------------
Is this a new environment or have you been able to perform dumps/loads against these instances in the past, and at the same speeds?
Hi Marco,
Pick the server who most likely was the old primary and issues the following commands:
Logon To ASE as user sapsa:
use master
go
sp_hadr_admin primary, ‘force’
go
sp_hadr_admin activate
go
Reagrds,
Joachim
Hi Maria
Are you sure DBCLIENT can be use for SYBASE ASE 15.7 ?
I am also in same problem
Hello Rahul,
First of all, as reminder, we are talking about ASE for Business Suite here. In such context, DBCLIENT package (v.15.7 or 16) is the only client you can use for ASE 15.7/ASE 16.0 for Business Suite.
I.e last available patch for ASE 15.7 is DBCLNT157SP17_0-20008025.SAR - Sybase ASE 15.7 FOR BUS. SUITE DBCLIENT.SAR SP17 PL0
You can find it in SWDC -> Databases -> SAP Adaptive Server Enterprise -> Database patches -> SAP ASE FOR BUSINESS SUITE -> SYBASE ASE 15.7 FOR BUS. SUITE
Hope this clarifies
Regards,
Cris
Hello Cesar,
Could you please post your question in the ASE for Custom Applications space. (http://scn.sap.com/community/ase-custom-applications) ?
This space is specific to SAP ASE for SAP Business suite (=> SAP Applications on SAP ASE).
Thank you
Regards,
Cristina Normand
SAP Active Global Support
Hello,
SCC is not specifically designed for use in an SAP Business Suite system. The recommended graphical database administration tool in an SAP Business Suite environment is the DBA Cockpit - transaction dbacockpit.
Regards,
Cristina Normand
SAP Active Global Support
Looks like you have processes that logg on when the DB is up.
Have you checked with sp_who again which users are on?
Maybe sceduled jobs?
Cron jobs?
In that case Dwayne is right - start the DB with option -m. That should help.
Hello Experts,
While i'm starting My instance in SAP MMC dispatcher halts . I figured out that the database of my instance was not up and that's why dispatcher halted. When i manually staring database from services ,Dispatcher turn green and everything is cool.But i want to know why this happening ..
Please enlighten me with some ideas?May be some missing parameter in startup profile?
Kind Regards,
Anoop