Item: TIBCO Integration (including BusinessWorks and Flogo)
Rating: 3
Author: Verified User

Use Cases and Deployment Scope

Our IT team is using Scribe to migrate data from one cloud application to another as part of a transition to a new application. It will also be used after the data migration is complete on a daily basis for regular data processing for interfaces between the cloud and our internal organization's application. We are the first team in our organization to try Scribe. We have used other services similar to Scribe, but they are more costly.

Pros and Cons

Scribe has a lot of connectors for different types of sources and targets
Security is easy to maintain in Scribe
Scribe's use of graphical mapping designer makes it easy to visualize what a map is doing

List view filtering is annoying in the interface. It does not save your filter or page that you were viewing. So if you click on a solution and then return to the list, the view is reset. You also cannot filter on solution status.
Often the metadata for sources and targets do not display in mappings. Sometimes you have to close and reopen a map several times before the fields will display. It's extremely slow.
There is no way to stop run-away processes via the application. I've had many instances where a process has been hung and cannot be stopped without calling in a ticket. Usually, it perpetually stay s in that state and you end up needing to recreate the solution.
There is no way to purge the internal database file used on the Scribe agent. Once it reaches a maximum size (which is small), the agent becomes unusable and needs to be retired.
It's difficult to manage which solutions are using which agents and which solutions are using which connections. There should be easier ways to report on this information.
Often your changes appear to have been saved but are actually not saved and the solution does not run correctly.
I've had instances where a source Oracle view is inputting x number of records but Scribe does not process them and does not produce any sort of error. It just looks as if there was nothing to process. This happens randomly.
The "statistics" produced by the solutions do not make it easy to understand which maps produced which update/create counts. Also, the record counts are difficult to interpret and also thrown off by batching. Other cloud services seem to be able to produce easy to read record counts by step even when they include batching. Not sure why this cannot.
There is no good instruction for determining the correct number of maps per solution or number of agents per server that are recommended.
There is no facility for running pre and post processing commands (like in other similiar apps). So you cannot have a script to check for presence of a file or to delete a file or create a trigger file.
Flat file handling is not very powerful and is clunky. If you want to change the delimiter used by a source or target file, you need to manually recreate the entire definition. You can import from a sample file, but the individual field type and sizes are lost.
If you want to reassign a connection for a solution (like UAT to PROD), you need to go into each individual map and change the source there instead of being able to do it at the solution level.
There is no way to copy a solution. You can have to export all the maps individually, create a new solution and then import all those maps. Then go into each one and wait for the metadata to refresh and then save each one (sometime several times) in order for it to become "complete".
The progress counter is very slow to update on running solutions
When a solution is running, there is no way to tell which map is currently executing within the solution.
If you run an individual step within a solution instead of the entire solution, there is no way to tell which step is running until it's done.
Validation within a map is cloogy. the slowness of the interface causes lag in the map validation. this flags errors that no longer exist.
Mapping fields could be easier. there is no marker on the source side to show which fields have been mapped and which haven't. This is a pain when you have hundreds of fields to map. Other similar apps show a check mark.
There are so many little things that could be improved about the user interface. I could go on and on.

Return on Investment

The data migration has been extremely slow. Things that should have taken a short amount of time have gone way past where any of us though.
Job monitoring has been more resource intensive and time consuming that planned for.
Researching and searching for things is time consuming.
There is no way to easily organize solutions. So there are hundreds of solutions in the list with no easy way to put them in folders or something like that. The only thing you can do is name them specifically so that you can filter them. But filtering is not very good in Scribe.
The concept of Scribe is good. I like the graphical flow builds in the maps. It just doesn't work very well yet. maybe as they improve it.

Business Processes Supported

Salesforce to Microsoft Dynamics CRM data integration
Oracle to Microsoft Dynamics CRM data integration

Users and Roles

10 - IT project team for initial data migration
IT support team for day to day feeds

Alternatives Considered

Informatica Cloud

Infromatica Cloud is much better and more stable, but also much more expensive. Price won out in the end.

Key Differentiators

Price
Product Features
Third-party Reviews

The product was recommended to us by the consulting company hired to manage the product since they had prior experience with Scribe. And the price was better than similar competitors.

Training Types Used

Self-taught

We did have several meetings with Scribe experts but I didn't find it that helpful. Mostly I have found things out by trial and error or searching internet forums to see if anyone else has had the issue. In most cases, the issues are known issues with no resolution or just something you cannot do with the tool. It would be nice to see more videos or help sites with examples that can get people started on things. Lack of examples seems to be the biggest lack.

TIBCO Cloud Integration Impact

Scribe performed almost exactly the same function as our prior ETL tool. The prior tool had a better interface (IMO) but Scribe is less expensive. Scribe didn't really offer any benefits other than cost and at the time had better integration with Microsoft Dynamics CRM since the other tool wasn't ready with the CRM connector.

Likelihood to Recommend

It's suited for low volume data movement on sources and targets which don't have many fields and are located on databases without many objects. The amount of metadata generated when retrieving database schemas causes extreme slowness when you have many fields and/or numerous DB objects. We tried to create schemas with access to as little as possible to limit this. It also should be something closely monitored if using it for scheduled tasks. It often fails for varying reasons, and needs to be restarted or have individual records reprocessed. It's not incredibly stable.
Because of the diverse connector types available, Scribe is useful for accessing a variety of source/target types.

Innovative Uses

none

Products Replaced

Support Headcount Required

3 - support team should have knowledge of the source and target connection products (ex. Oracle, Microsoft Dynamics CRM). The Scribe tool itself can be used by anybody with logical thinking ability and familiarization with programming concepts in general.

Future Planned Uses

Salesforce.com
Flat file integration with Oracle

Likelihood to Renew

The project which required the Scribe tool has been cancelled due to business direction change. Unless another project requiring Scribe comes up, it will likely not be renewed. Companies are always looking for cost savings though and Scribe is less expensive than competitors, so you never know.

A lot of potential, but it's got a long way to go

Overall Satisfaction with TIBCO Cloud Integration (including BusinessWorks and Scribe)