You have also been frustrated over having to find a document for the business, so they could see what kind of messages they had processed. Having to open SXI_MONITOR and find the relevant documents.

PIArchiving is a tool that makes it easy to the original messages. It can for instance be when dealing with partners and the invoices, and make sure the invoices can be found in legal situations. PIArchiving is focus on making a solution that is easy for the business to use to find the documents they want.

For this release I have been working a lot with improving the performance and stability. And in the process I also removed some bugs. PIArchiving is now much better to use in your productive systems and easier to get started with.

New parsing format added.
I also added MT940 parsing of the messages, so the business easy can find the original documents based on bank statement number and bank account. Just imagine how much easier it is if the finance department can monitor the messages they have received. It is still possible to parse XML, EDIFACT and X12 formats and extract relevant data from them.

Performance test
I have done a performance test to find how well the solution is performing. It is costing some extra performance to parse the documents and save the documents as expected. The impact of this extra parsing does not affect the sizing of your PI system. Check out the results on this blog post on the PI performance.

I have kept the simplicity so you should be able to start using the service within a few hours. And there is a free 30 day trial, so it is really simple to get started.

The pricing of the solution is €3997 so it is probably one of the world’s cheapest SAP solutions.

I had setup a really simple scenario, where I was reading a 10.000 18kb XML file from a local ftp server. I had just created a sender agreement to the file stopped with errors due to missing receiver determinations. All messages therefore only had limited processing time in PI and only took time for the adapter performance. The system was processing 1262 messages per minute.

Then I added SAP PI archiving and parsed the XML and run 3 different XPaths. The data is saved in the database and the file is archived in the local file system. In this mode the system could read 558 Messages per minute. This is the most complex processing because the system needs to build an XML document in memory and evaluate the XPaths. The EDI parser should be performing much better.

The system was a SAP PI 7.1 running on an AIX system with an unknown size. It is therefore only possible to see the data as relative.

In a real scenario there is a mapping and another adapter communication. So this will lower the default performance and mean the overhead of the PI Archiving is not that large in comparison.

All of the process in the setup was only focus on one PI communication channel. There was therefore no process working in parallel. It would be possible to have more processes on multiply channels.

It does cost some performance to use the PI Archiving but it is not something that will make it much different regarding the performance in normal situations. Batch imports are processed still with 558 messages a minute and it should be enough.

Try SAP PIArchiving now, download it now for free.

For the PIArchiving solution it is of key importance to have a cheap easy way to store messages for a long time. AWS S3 provides this making it easy for users to store their data in the systems. Data is stored securely in S3, by being encrypted when it is uploaded.

PIArchiving also uses SimpleDB to store metadata about messages. It makes it possible to find the invoice with the number 123 or the message sent from partner ABC. There is parsers for EDIFACT, X12 and XML.

See the website http://aws.amazon.com/solutions/solution-providers/piarchiving/ for more information on the SAP PI Archiving solution.

I see that there are two different types of archiving

• Technical
• Content archiving

Technical

This part is where messages is saved technical. It is the standard form of archiving in SAP Process Integration. Technical archiving is where all the runtime messages from SXI_MONITOR and the corresponding tables from the adapter framework are removed. This happens to conserve space in the database and ensure the PI can run fast enough.

Content

This part is to make sure the content sent is archived. This is by extracting the message and make sure it can be recovered.

There is a 15 minutes video about the content.

One of the things that really bothered me when working with PI was that it was to difficult to find the messages that has been processed. I had to download the files from the FTP server and to a text search to find the document with invoice 1234. Or even worse I should go thru SXI_MONITOR and open all messages in the timeframe to find the correct message.

I also had some issues with saving the data on a FTP server that filled up. This stopped the SAP PI system which was not really an excellent idea.

Those were the main reasons that I decided to develop my SAP PIArchiving solution. The focus has been on making a solution that is easy to use and does not require a lot of configuration.

The initial ideas were just to be able to upload messages to Amazon S3, because it was simple and quite reliable as a place to store files. Several people told that they was not really happy with storing their data externally. So the solution had to be refactored to also allow local based storage. Currently the solution supports the local file system, but other solutions are going to be built like FTP(just make sure you have enough space ) and the SAP content server. It is fairly easy to add more functionalities to the solution.

I created a short video to show how simple the solution is.

Please leave a comment on what you think or come with ideas for improving the solution.