This project has moved and is read-only. For the latest updates, please go here.

Historian 2.0 Information

Dec 17, 2012 at 8:11 PM

I haven't heard any new information on the Historian 2.0... actually, I didn't attend the forums, so I don't know what was expected. Is there any information, spec sheets, possible release dates, etc? I am developing software for data quality, and don't want to duplicate your efforts.

Dec 17, 2012 at 9:52 PM

I’m glad you asked. We have been very busy these past few weeks putting the finishing touches on the individual components that make up the historian. The final step is to wrap all of this functionality into a user friendly API layer. We are planning a beta release as soon as we feel the API layer is well defined.

If you would like to take a look at the alpha code, you can visit http://openhistorian.codeplex.com to download it.

We have been very impressed with the performance we have seen in benchmark testing. While benchmark testing does not always represent real world performance, we are seeing write speeds at approximately 1 million points per second and sequential read speeds in the 10 million points per second range.

We are also seeing a significant reduction in space requirements over the current version of the historian 1.0. Anywhere from 50% to 75% less disk space has been observed when compressing (lossless) a historian 1.0 file.

If you are interested in developing against the historian 2.0, the historian’s fundamental data storage is as follows: a 64-bit time field; a 64-bit id field; a 64-bit user definable quality field; and a 64-bit user definable value field.

The best way to test the historian is in a real world environment. OG&E has volunteered to be our test space. We are currently break/fixing the code in this environment. Therefore when it is released, it will have at least one production environment that it is operating in.

Our release schedule is planned for early 2013. Be looking for a beta release sometime in January.

Dec 18, 2012 at 4:33 PM
thanks for the response. the improved storage is important, but for my needs any enhancements to the UI and/or the API are of the greatest interest.


On Mon, Dec 17, 2012 at 4:52 PM, stevenchisholm <notifications@codeplex.com> wrote:

From: stevenchisholm

I’m glad you asked. We have been very busy these past few weeks putting the finishing touches on the individual components that make up the historian. The final step is to wrap all of this functionality into a user friendly API layer. We are planning a beta release as soon as we feel the API layer is well defined.

If you would like to take a look at the alpha code, you can visit http://openhistorian.codeplex.com to download it.

We have been very impressed with the performance we have seen in benchmark testing. While benchmark testing does not always represent real world performance, we are seeing write speeds at approximately 1 million points per second and sequential read speeds in the 10 million points per second range.

We are also seeing a significant reduction in space requirements over the current version of the historian 1.0. Anywhere from 50% to 75% less disk space has been observed when compressing (lossless) a historian 1.0 file.

If you are interested in developing against the historian 2.0, the historian’s fundamental data storage is as follows: a 64-bit time field; a 64-bit id field; a 64-bit user definable quality field; and a 64-bit user definable value field.

The best way to test the historian is in a real world environment. OG&E has volunteered to be our test space. We are currently break/fixing the code in this environment. Therefore when it is released, it will have at least one production environment that it is operating in.

Our release schedule is planned for early 2013. Be looking for a beta release sometime in January.

Read the full discussion online.

To add a post to this discussion, reply to this email (openpdc@discussions.codeplex.com)

To start a new discussion for this project, email openpdc@discussions.codeplex.com

You are receiving this email because you subscribed to this discussion on CodePlex. You can unsubscribe or change your settings on codePlex.com.

Please note: Images and attachments will be removed from emails. Any posts to this discussion will also be available online at codeplex.com




--
Patrick Pentz
Dec 18, 2012 at 4:42 PM

We have a primitive API right now. I'm currently working on better documenting it so we can post some examples to the website. If there is a way you would like the historian presented, we can always write wrapper APIs to meet your needs.

Dec 18, 2012 at 4:53 PM
The need is to supply historian data to various report writers, even Excel. I imagine the current CSV output can be used for such purposes.


On Tue, Dec 18, 2012 at 11:42 AM, stevenchisholm <notifications@codeplex.com> wrote:

From: stevenchisholm

We have a primitive API right now. I'm currently working on better documenting it so we can post some examples to the website. If there is a way you would like the historian presented, we can always write wrapper APIs to meet your needs.

Read the full discussion online.

To add a post to this discussion, reply to this email (openpdc@discussions.codeplex.com)

To start a new discussion for this project, email openpdc@discussions.codeplex.com

You are receiving this email because you subscribed to this discussion on CodePlex. You can unsubscribe or change your settings on codePlex.com.

Please note: Images and attachments will be removed from emails. Any posts to this discussion will also be available online at codeplex.com




--
Patrick Pentz
Dec 18, 2012 at 5:22 PM
Many report writers (and Excel) can access data via ODBC. So if any API is created for the historian, an ODBC driver would be my first desire.


On Tue, Dec 18, 2012 at 11:53 AM, Patrick Pentz <patpentz@gmail.com> wrote:
The need is to supply historian data to various report writers, even Excel. I imagine the current CSV output can be used for such purposes.


On Tue, Dec 18, 2012 at 11:42 AM, stevenchisholm <notifications@codeplex.com> wrote:

From: stevenchisholm

We have a primitive API right now. I'm currently working on better documenting it so we can post some examples to the website. If there is a way you would like the historian presented, we can always write wrapper APIs to meet your needs.

Read the full discussion online.

To add a post to this discussion, reply to this email (openpdc@discussions.codeplex.com)

To start a new discussion for this project, email openpdc@discussions.codeplex.com

You are receiving this email because you subscribed to this discussion on CodePlex. You can unsubscribe or change your settings on codePlex.com.

Please note: Images and attachments will be removed from emails. Any posts to this discussion will also be available online at codeplex.com




--
Patrick Pentz



--
Patrick Pentz
Jan 10, 2013 at 7:53 PM

Our group is really interested in Open Historian project as well, so excuse me for using this post to ask my question. I know we can extract data from Historian using HistorianPlayBackUtility. But is there anyway that Matlab can access Historian directly? We use Matlab to do data analysis and visualization a lot. Just curious. Thank you!

Jan 10, 2013 at 8:16 PM

I'm not very familiar with MatLab. I used it a little in college, but not enough to know what are the requirements for importing data via code. The openHistorian 2.0 will have many ways to programmatically extract data from a data source. We are also considering having some of the more popular database driver methods in the base package. However, we understand that we may have to rely on the open source community to get this product extended where applications from all types can take advantage of this superior historian architecture.

If you have any specific details about what type of data you plan on storing/analyzing, it may help us make sure that our solution is well architected. Feel free to contact me through codeplex's site on any details.

Jan 11, 2013 at 4:24 PM

Steven,

Thank you so much for your reply! I am really looking forward to openHistorian2.0!

Right now, we store like 4GB measurement data per day in Access DB, which really needs to change. So we are eager to see whether Historian can be our future data storage solution.  From what I understanding, Matlab can handle odbc and jdbc driver. So it can actually import data directly from Access, but it is extremely slow. Currently, we always extract .csv files from Access DB(using a c++ program) and then let Matlab to import them. I am just curious whether we can make everything completely in Matlab if we use Historian

Another thing we concern is how easy or how fast it would be to transfer our 8 years of Access database to Historian. 

I completely understand that we need community input to a lot of these requirements.

Thanks again!