Information Server File Formats

Unlike traditional programming languages the source code of DataStage assets is expressed using a set of proprietary export formats. DataStage assets can currently be exported in three possible formats:

1. DSX – The older Information Server export format that supports a subset of all current Information Server asset types.

   1. This format is also available 7-bit encoding in which any ASCII character above DEL (127) is encoded as \nnn where nnn is the relevant character’s ASCII number.

2. XML (referred to in the DataStage documentation as ‘Legacy XML')

3. ISX – A compressed XML (binary) based format which supports all Information Server asset types.

To add further complexity to this picture the ISX format exists in multiple variations:

• The ‘default’ format, useable by all IBM tools to represent single DataStage jobs

• An Information Server Manager-­specific ISX format.

• An ISTool releases (multiple job versions)

For DataStage Jobs, the DSX and ISX formats allow you to choose whether you want your export to include Job design information only, encoded-binary executable information, or both. When you choose to create an export incorporating only Job design information then that file will need to be re-compiled in any target project into which it is subsequently imported.

Each format occupies varying amounts of disk space, with the design-only ISX format occupying the least. By way of example here is a set of exports (of a simple 3-stage Parallel Job) using various formats, order by increasing file size:

You’ll see that the ISX design-only format occupies the least space.

Why does MettleCI use DSX files?

When performing a Git commit MettleCI uses the ‘default’ flavour of the ISX format with only design information. Executable information is deliberately excluded. (I’ll link to the “But how do we do binary-only deployments to PROD?” FAQ here)

MettleCI provides the mechanisms for committing DataStage artefacts into a Git repository, testing them against Compliance Rules and Unit Tests, and deploying them to downstream environments, including the swapping of Jobs' parameter values to those appropriate to each target environment. All of these operations relay on the capabilities of the ISX format.

The format in which DataStage artefacts are stored in Git only matters to a customer when they want to do something with

Does the re-compilation of jobs in each environment compromise good deployment governance

File Contents

Treat the export as opaque

While the DSX format might be text based like traditional source code, both DSX and ISX formats represent each DataStage job as an acyclic graph with complex relationships and extensive metadata properties. The graph-based nature of these files means that even though Git might be able to merge text based DSX files, the resulting output will contain a large number of both textual and semantic conflicts. To make matters worse, a developer could open the exported files in a text editor but, since these are not intended as human readable, it is virtually impossible to fully understand the DataStage asset they represent, making conflict resolution near impossible. For all practical purposes DataStage exports should be considered as binary files within the context of a version. control system.

Git will still be able to perform a merge, but if two different versions of the same DataStage export are detected Git will report a conflict and the versions will need to be manually merged. Since DataStage does not include any tools for merging two DataStage exports the only way to resolve a conflict is for a developer to ‘eye-ball’ each version in the DataStage user interface and manually construct the merged version. This process is time consuming, tedious, and error prone.

A desire to use DSXs stems from:

The perception that ISX files are opaque binary files to which existing build and deployment processes and can include the compiled version of the job.

We use a custom DSX processing script to split DSXs or to swap parameter values appropriate to the environments into which they’re being imported.

We want to be able to DIFF DSXs to compare versions in version control, or between environments. WE can’t do that with ISXs

MettleCI’s use of design only assets Some customers have security constraints meaning that only executable assets can be

Conclusions

MettleCI means that the file format used to store your jobs in Git version control doesn’t matter. This is because:

Your build and deployment process won’t be relying on custom text-processing scripts to modify the controlled assets.

You won’t be relying on traditional test-based diffing techniques to identify the changes between code versions.

MettleCI uses the 'default' ISX format (1) which contains a single version of the job which can be exported and imported without imposing restrictions about the tooling which ca be used.

The DSX format (regardless of encoding) is a proprietary ASCII-based format.

The ISX format

wrapped XML are ASCII readable text, they are not easily 'human readable', and there is little to no value inspecting a DataStage Job's source code outside of the DataStage designer itself.

DataStage Jobs cannot be loaded into an external diff tool to identify the changes between them, and DataStage job exports contain stages in a non­deterministic order, meaning that two successive exports of an unaltered job may mistakenly identify significant change when compared using a traditional diff tool. The only way to read and understand a job or compare differences between two jobs is to let DataStage load them into memory and present a logical representation which is human readable. This process remains the same regardless of whether the export is encoded using binary or ASCII.

There are a number of good reasons to select ISX as the management format for Information Server artefacts:

• ISX's are compressed, and Git only stores deltas so the disk footprint is minimised

• Git will attempt a merge on non-­binary files, which would likely result in the corruption of the asset

• ISX is the only format which has support for all Information Server asset types (not just those in DataStage) so is more useful

Both ISX and DSX formats include a lot more information that changing update, export and compile dates. They also contain view port position, zoom levels and snap to grid. They also contain a lot of "non­functional" information such as link label positions. Normalising the export might cut down some noise but does not provide a robust way to determining if a given job has changed between check­ins.

Earlier iterations of MettleCI did not provide the quick and easy check­in process that we have already demonstrated. Checking in jobs manually was tedious and time consuming, so check­ins were usually performed in bulk. Since there was so much time between a change being made and a developer performing the check­in, there were a high amount of "false" check­ins. We did use normalisation to reduce some of the noise but it wasn't robust enough to identify all unchanged jobs.

Since introducing the MettleCI check­in process, it is very rare for developers to check­in a job even though they haven't made any changes. On the odd occasion that it does occur, there isn't really any impact ­ the job will be retested in the CI build and a little more space (kilobytes worth) will be used up by Git storage.

If, for some reason, developers are regularly unsure of what has changed, then the team should question why there is such a long gap between modifying a job and checking it in.

JMcC:

ISX file size wouldn’t be a valid reason to object. The differences between the design-time ISX and DSX export formats are negligible with respect to per-repository storage limits for Git server platforms like Github. If they start using MettleCI’s Unit Test feature, the per-test data files (input and expect output) will likely eclipse the storage footprint of their Jobs before long, anyway. (edited)

Justin McCamish:speech_balloon: 4 hours ago

Where we’ve encountered initial anxiety about adopting ISXs in the past, it usually boils down to either:

Aspects of the customer’s current Job wrangling process relying on the parsing of DSX contents (whether programmatically or manually). MettleCI usually renders that redundant but unpicking entrenched thinking becomes the key challenge (i.e. change management-101).

A misunderstanding about ISX files being intrinsically “binary” as opposed to containing a less-human-readable-but-more-detailed representation of a Job design than DSX. As you know, both formats offer the optional inclusion of a compiled binary payload but MettleCI doesn’t need or use that in any its functions.

(edited)