Following is excerpt from the tnsnames.ora file:
SUMSKYDB =
(DESCRIPTION =
(ADDRESS_LIST =
(LOAD_BALANCE=ON)
(FAILOVER=ON)
(ADDRESS=(PROTOCOL=TCP)(HOST=host1-oracle-vip)(PORT=1521))
(ADDRESS=(PROTOCOL=TCP)(HOST=host2-oracle-vip)(PORT=1521))
)
(CONNECT_DATA =
(SERVICE_NAME = SUMSKYDB.SUMMERSKYUS.COM)
(FAILOVER_MODE = (TYPE=SELECT)(METHOD=BASIC)
)
)
)


It is understood that, when one Oracle instance fails, it will
transparently connect to the other instance.
I would like to know about, to what instance the preference will be
given to when both the instances are up and running. Is this according
to the order that is given in tnsnames.ora file? In the above example,
it will always be to 'host1-oracle-vip' first and then to
'host2-oracle-vip' in case if 'host1-oracle-vip' is not running. Or,
some other factor decides it?

Could anybody clarify?

Hi,

When you have TAF configured without load balance at the client side,
it will always go the first connection. If it is not able to connect
the first named connection, it will go to the subsequent connection. If
you specify the load balance it just uses the round robin method and
choose the connections alternatively. Remember TAF is the feature from
the network layer, not servics layer.

On 22 May 2005 08:30:03 -0700, qazmlp1209 (AT) rediffmail (DOT) com wrote:

First come first served.

Please do not crosspost

--
Sybrand Bakker, Senior Oracle DBA

K Gopalakrishnan wrote:
Minor correction. Oracle does not round robin when you ask for load
balancing. The connections are smart enough to know which machines are
burdened and to take into account which machines have faster CPU,
faster bus, etc.

I have seen many times when the first 10+ connections go to a single
node before the balancing sees any reason to move subsequent connections
to other nodes.
--
Daniel A. Morgan
http://www.psoug.org
damorgan@x.washington.edu
(replace x with u to respond)

Daniel:

I think there was mis-understanding as I have not discussed the load
balancing to the depth as the original poster just wanted to know from
the tnsnames.ora level. That is why I had clarified him saying TAF is a
feature from the network layer, not from the services layer and TNS
load balancing is purely round robin. It does not care about CPU load
or number of connections.

To answer your question, you are commenting about listener load
balancing where we register the CPU load to the other nodes (using
remote listener) and do the load balancing based on the least loaded
note. And also there is a way to disable that using an underscore
parametetr (_prefer_least_loaded_node or something like that in
listener.ora) as the listener load balancing does not perform
effectively for benchmarks or logon storms.

If you have the listener load balancing setup, the listeners cross
register with other participating instances and update the cpu load in
every 5 minutes. If you have all connections coming up with in 5
minutes there are chances that all connections will go to the first
identified node. I think that is what you have experiened and that is
not based on numbers (10+ connections as you mention), but based on the
CPU load during that time.

There are couple of other load balancing options also available based
on other network level parameters. But to answer your question,
listener load balancing is based on machine load and client load
balancing is based on round robin.

Hope this clarifies this

Best Regards,
K Gopalakrishnan
Co-Author: Oracle Wait Interface, Oracle Press 2004
http://www.amazon.com/exec/obidos/tg.../-/007222729X/

K Gopalakrishnan wrote:
I didn't know I was unclear about anything ... but thank you for
clearing it up for the many others that may have been wondering
what this was all about.
--
Daniel A. Morgan
http://www.psoug.org
damorgan@x.washington.edu
(replace x with u to respond)