How to define Sybase DataSource in IBM WebSphere Application Server

[philip_yang]

I have a Sybase server (Adaptive Server Enterprise V12)
configured as High Availability. The connection properties
are stored in two LDAP servers.

Currently, I am using
DriverManager.getConnection(db_url, connectionProperties) to
create database connections. And db_url is in the following
format:
jdbc:sybase:indi:ldap//LDAP_IP:port/servername=DBServer,searchBase.

To leverage the WebSphere's connection pool feature, I'd
like to define a DataSource for my server. However, I'm not
sure if WebShpere is smart enough to know that the
configuration properties are kept on the LDAP server. Please
help.

[John MacEnri]

Sorry, forgot to add what the resultant URL looks like

jdbc:sybase:Tds:<host>:<port>/<dbname>?REQUEST_HA_SESSION=true&SECONDARY_SER
VER_HOSTPORT=<ha_host>:<ha_port>

Regards

John

"John MacEnri" <jmac...@imsmaxims.com> wrote in message
news:3f7acd72$1@forums-2-dub...
> Hi,
>
> I was recently configuring a Java Application to connect to a HA cluster
of
> 2 ASE's. I went the way of setting up the LDAP server and all that, but
then
> found it's actually not required in later EBF's of the JCOnnect 5.5
> I am using EBF 11473 which is very recent, but I think the change has been
> in for about 3 EBF's now.
>
> They now support a much simpler way of configuring the details of the
> secondary fail-over ASE.
> All you need to do is set 2 connection properties.
> REQUEST_HA_SESSION=true
> and
> SECONDARY_SERVER_HOSTPORT=<host>:<port>
>
> where host:port is the details of the fail-over server.
>
> With these connection properties set, every connection created in the
> connection pool, will have automatic connection fail-over enabled.
> The only problem you're left with is you have to code for the exception
that
> is thrown when the fail-over occurs (JZ0F2), if you want the current
> transaction automcatically re-tried.
>
> Regards
>
> John
>
> <Philip Yang> wrote in message news:3f706608.60...@sybase.com...

[John MacEnri]

Hi,

I was recently configuring a Java Application to connect to a HA cluster of
2 ASE's. I went the way of setting up the LDAP server and all that, but then
found it's actually not required in later EBF's of the JCOnnect 5.5
I am using EBF 11473 which is very recent, but I think the change has been
in for about 3 EBF's now.

They now support a much simpler way of configuring the details of the
secondary fail-over ASE.
All you need to do is set 2 connection properties.
REQUEST_HA_SESSION=true
and
SECONDARY_SERVER_HOSTPORT=<host>:<port>

where host:port is the details of the fail-over server.

With these connection properties set, every connection created in the
connection pool, will have automatic connection fail-over enabled.
The only problem you're left with is you have to code for the exception that
is thrown when the fail-over occurs (JZ0F2), if you want the current
transaction automcatically re-tried.

Regards

John

<Philip Yang> wrote in message news:3f706608.60...@sybase.com...

[lance.a...@sun.com]

Looks like you are not using a DataSource, but a DataSource
wrapper as for a DataSource you will not specify a URL, this
is only for DriverManager

> Sorry, forgot to add what the resultant URL looks like
>
> jdbc:sybase:Tds:<host>:<port>/<dbname>?REQUEST_HA_SESSION=

> true&SECONDARY_SER VER_HOSTPORT=<ha_host>:<ha_port>