DataDeploy Design Question

Hi Devnetters -
I've used OpenDeploy for several years now (great product, imho), but never DataDeploy.

I met with some potential clients yesterday that had the following scenario:

They have an admin tool that produces both web content (files) and associated data (in Oracle). The web content portion of the output can easily be housed in our TeamSite repository.

They'd like to, when a publish event happens, have a workflow pick up the newly created web content and move it to another server. This is standard workflow development.

Additionally, they'd like the same workflow / deployment process to extract the rows (that are associated with the web content) from the source Oracle db and sync it to a target Oracle db. This is where I'm not clear if DataDeploy can help or not.

At a design level, I'd like OpenDeploy / DataDeploy to execute a query against the source db, package up the results (into a dml statement?), transport the results package to the target machine and apply the update to the db.

I've scanned the DataDeploy and OpenDeploy manuals for any scenarios similar to this and I think I'm a bit more confused than when I started! Can any of you folks with DataDeploy experience give me some advice on if and how this might be done? Any help is much appreciated.

Thanks,
Wally Box
Nike, Inc.

Find more posts tagged with

Comments

Adam Stoller

Hi Devnetters -
I've used OpenDeploy for several years now (great product, imho), but never DataDeploy.

I met with some potential clients yesterday that had the following scenario:

They have an admin tool that produces both web content (files) and associated data (in Oracle). The web content portion of the output can easily be housed in our TeamSite repository.

They'd like to, when a publish event happens, have a workflow pick up the newly created web content and move it to another server. This is standard workflow development.

Additionally, they'd like the same workflow / deployment process to extract the rows (that are associated with the web content) from the source Oracle db and sync it to a target Oracle db. This is where I'm not clear if DataDeploy can help or not.

At a design level, I'd like OpenDeploy / DataDeploy to execute a query against the source db, package up the results (into a dml statement?), transport the results package to the target machine and apply the update to the db.

I've scanned the DataDeploy and OpenDeploy manuals for any scenarios similar to this and I think I'm a bit more confused than when I started! Can any of you folks with DataDeploy experience give me some advice on if and how this might be done? Any help is much appreciated.

Thanks,
Wally Box
Nike, Inc.

If you're doing Oracle-to-Oracle deployments - you should look for solutions from Oracle for doing it ... DD is not intended for DB-to-DB deployments.

Alternatively, if you cannot find an Oracle solution - I'd recommend looking at setting up Perl DBI (or Java equivalent) rather than trying to hack DD to do what you want.

Migrateduser

Thanks Ghoti -
Since I posted this, I did find the Use Case matrix in the DataDeploy manual where it specifically says that DataDeploy is not intended for database-to-database replication.

If I step back from any particular tool selection and look at the workflow requirement I'm trying to fulfill, what I'm trying to do is to package up both content and data into a transaction that, when applied to the target server, leaves the target server in sync with the source server.

Then, when I move forward a bit into assuming that I'm using the TeamSite / OpenDeploy combo as the pipeline to move the assets from the source to the target, I keep coming back to looking at whether DataDeploy can be used to apply the update to the data portion of the target.

In that same way that OpenDeploy is TeamSite 'aware', I would hope that DataDeploy is OpenDeploy 'aware' and that that awareness could be leveraged into making the transactional content / data update happen - using tools from one vendor, and controlled by the workflow with a minimum of customization.

Now, if I were asking about the same scenario and the data on the source was stored in the content's metadata in the TeamSite repository, then would DataDeploy be the 'correct' solution?

And, taking that a bit further, if I were to have an xml-based extract of the data to be sent, would DataDeploy be a good choice (as it will take an xml file as input)?

I'd rather some work was done up front on the source server to get the data into a format that the workflow / OpenDeploy / DataDeploy combo would work with than integrate other toolsets in - but then I haven't done this before and is why I'm asking here.

Comments?

Thanks,
Wally

Adam Stoller

Thanks Ghoti -
Since I posted this, I did find the Use Case matrix in the DataDeploy manual where it specifically says that DataDeploy is not intended for database-to-database replication.

If I step back from any particular tool selection and look at the workflow requirement I'm trying to fulfill, what I'm trying to do is to package up both content and data into a transaction that, when applied to the target server, leaves the target server in sync with the source server.

Then, when I move forward a bit into assuming that I'm using the TeamSite / OpenDeploy combo as the pipeline to move the assets from the source to the target, I keep coming back to looking at whether DataDeploy can be used to apply the update to the data portion of the target.

In that same way that OpenDeploy is TeamSite 'aware', I would hope that DataDeploy is OpenDeploy 'aware' and that that awareness could be leveraged into making the transactional content / data update happen - using tools from one vendor, and controlled by the workflow with a minimum of customization.

Now, if I were asking about the same scenario and the data on the source was stored in the content's metadata in the TeamSite repository, then would DataDeploy be the 'correct' solution?

And, taking that a bit further, if I were to have an xml-based extract of the data to be sent, would DataDeploy be a good choice (as it will take an xml file as input)?

I'd rather some work was done up front on the source server to get the data into a format that the workflow / OpenDeploy / DataDeploy combo would work with than integrate other toolsets in - but then I haven't done this before and is why I'm asking here.

Comments?

Thanks,
Wally

If you package up the DB extraction and DB deployment (via Perl DBI or equivalent) then you can make sure that the 2nd part of that operation succeeds as part of an OD DNR script - which will give you the transactional aspect you're looking for.

Unless there's a reason to version the data collected from one DB before it gets sent to the other DB - I wouldn't bother trying to bring it into TeamSite.

If there is - you can still go the DBI route for the extraction part and then you can decide whether you want to use DBI or DD to send it to the target DB - but the same philosophy applies - it's a DNR script and as long as you verify that the deployment to the DB succeeds, you get your transactional behavior.

ERROR.jpg