Latest version: 1.0.5

Sideview Process Historian

If you use OSISoft's PI System, the app comes with a Windows service that makes collecting the data a snap. Within minutes you'll have the service connecting directly to OSISoft's PI SDK and forwarding the data to Splunk. If you don't use the PI System contact us and we'll show you other ways to get your live data coming in.   Next »
The only other setup step is about tag naming schemes. If you use a naming scheme in your tags the app can analyze and aggregate your data more powerfully.   Next »
And you're off. The browse interface allows you to navigate the layers of your naming scheme. You are shown the most recent value for each tag value and you can flip over into the 'report' mode at any time.   Next »
When you browse down to the leaf levels the aggregation will stop, each cell will represent a single tag, and clicking the cell will lead you to run reports about that tag.   Next »
You can run simple reports over time, save them and share them with team members.   Next »
Or you can choose other values for the X-axis. No matter what kind of chart you generate, each column and legend item and line is always clickable and leads you to yet another 'drilldown' report.   Next »

App Details

Latest version: 1.0.5
First released: August 22nd 2011

Turn Splunk into a world-class historian for monitoring and reporting on your Scada process data.

This highly interactive app allows you to capture and report on manufacturing data in Splunk. Take the incredible scaling power and flexible reporting that Splunk originally developed for enterprise IT applications and use it to meet the challenges you face today monitoring complex manufacturing and utility systems.

This app comes with a Windows Service called “PI To Splunk”. If you are using OSISoft’s PI system and you have a server on which to install Splunk itself, setting up everything should take about 20 minutes. Even if you are not using the PI System on site, all is not lost. Let us know what you are using and we can advise you on short term ways to get your data indexed, and we can develop something more robust for your situation in future versions. Last but not least, for people who just want to take a look, the app comes with a simple script that can continuously generate sample data.

Download the beta, take it for a test drive and let us know what you think and what you would like to see in the future.

Splunk has taken the IT world by storm. It allows teams to analyze their data, run reports even at massive scale and use that insight to operate intelligently every day, no matter what’s going on. The Sideview Process Historian for the first time brings this revolution into the world of process manufacturing, utilities and beyond.

  1. Move your teams to a web-based solution without having to integrate with Sharepoint. You can set up Splunk to use Active Directory or any LDAP system for authentication, or you can just use Splunk’s own internal authentication system. Either way you can assign users to roles to limit their control or visibility as necessary.
  2. Enter a world where you can click on every element in every chart and see the data behind it. Every legend item, every pie wedge, every cell in every table is interactive and leads to another reporting interface. If users of your current historian and trender interfaces have been saying that drilldowns are cumbersome or nonexistent, download this app and see your world the other way.
  3. Over the last several years Splunk has radically evolved from a simple log search engine to a powerful data analytics platform. Beyond familiar statistics and timecharts on the values in your tag data, you can write searches and reports to look for broader patterns or catch systemic anomalies. Stand on the shoulders of giants by adopting technology developed for enterprise IT systems and start understanding root cause in your manufacturing systems today.

TAKE NOTE: while you are using the trial version, it will send some tracking information to Specifically it will send the name of the app, the name of the page you are loading within the app, the splunk version number as well as the app version number. If you would like to obtain a copy of the trial version without any phone-home functionality just send us an email.

Installation Overview

The Sideview Process Historian is a ‘Splunk app’ that you install within the Splunk Server, but it also includes a Windows service called “PI To Splunk” which must be installed separately on a server running OSISoft’s PI SDK. (If you are an OSISoft customer and you use the PI System but not the PI SDK, you can get the SDK at no cost from OSISoft from their Download Center.)

The PI To Splunk service will forward data to Splunk over TCP. Although you could theoretically put everything on the same server, we recommend installing Splunk on a separate server.

Installation consists of three steps.

  1. installing the Splunk Server.
  2. installing the Sideview Process Historian App within Splunk.
  3. installing the PI To Splunk service (comes bundled with the Sideview Process Historian App).

Installing the Splunk Server

We advise reading up about Splunk’s system requirements before proceeding. Be aware that between the ‘minimum’ specs and the ‘recommended’ specs, you will see a very significant difference in search performance even during trial usage. If you install Splunk on hardware matching the ‘recommended’ specs you can expect all searches and reports, and thus the user interface, to be roughly an order of magnitude faster than if you had gone with the minimum. Also think carefully before installing Splunk on a VM. Unless your environment can supply its VM’s with disk I/O performance numbers equivalent to a dedicated RAID 0, this is probably not a great idea.

All other details about installing Splunk can be found in Splunk’s Installation Manual

Installing the Sideview Process Historian App within Splunk

After you have installed Splunk, follow the instructions and log into its web interface. On the first screen after you log in, click the button in the upper right that says ‘Find more apps’. On the resulting screen search for “Sideview Process Historian” using the search bar at the top right. When you see the Sideview Process Historian App listed below, click the green button there that will say either ‘install’ or ‘download’. Follow the instructions on screen.

Once the app is installed, log in as the admin user, select the Sideview Process Historian App from the menu, and you should get taken to the Guided Setup page. If somehow you are not taken there first, look in the leftnav for the link that says “Rerun Guided Setup”, and click that. From there just follow the instructions on screen.

Installing the PI To Splunk service

Within the Sideview Process Historian you will see a top-level folder called “PI To Splunk”. Copy this folder to the server where the PI SDK is installed, and to a target location on that server such as “C:PI To Splunk”. This folder contains two subfolders, “PIDataWriter” and “PIDataWriter Config”. With the files in place, open the “PIDataWriter Config” subfolder and run PIDataWriterConfig.exe as Administrator (the easiest way is to right click PIDataWriterConfig.exe and select ‘Run As administrator’).

At the top of the PIDataWriterConfig interface, at the top left under ‘Service Control’, you will see an “Install” button. Click this first. This will bring up an installation dialog asking you to identify the location of the PIDataWriter.exe binary. Navigate to it, select the exe and click OK. Note for the future there is also a ‘Remove’ button that works the same way to uninstall the service.

After the service has installed, the PIDataWriter location should be specified again in the PIDataWriterConfig interface by clicking the “…” button and navigating to PiDataWriter.exe. After you have done this, various configuration options can now be applied to the service. These options are described below:

  • Refresh Rate — This tells the PI to Splunk service how fast to check the snapshot queue for data for each tag. A higher value means greater latency in sending data to the Splunk server. A lower value will reduce latency, but will increase the CPU load on the PI server and PI to Splunk service. Note that ultimately the frequency of the incoming data also depends on the refresh rate of the tags themselves within the PI System. If you set the Refresh Rate to lower than this intrinsic rate, it may well have no practical effect.
  • Tag List — This list specifies the list of tags that should send data to Splunk using a \servertagname format. Generally, this list is most easily populated using the tag search option by clicking on the tag search button. This will bring up the PI SDK tag search dialog, and allows common search options available in PI clients.
  • Splunk Server — The Splunk server name must be specified, and is commonly either a hostname of the Splunk server or a FQDN.
  • Port — The port specifies the Splunk listening port for socket communication. By convention the Sideview Process Historian App listens for its data on port 9993 on the Splunk Server. NOTE: If you decide to change this port from the default value of 9993 you will also have to make the corresponding change in the Splunk Server. To do so, navigate from the Splunk Interface to “Manager > Data Inputs > TCP”. You will see a row listing ‘TCP port’ 9993 (Note that you want the ‘Data Inputs’ section rather than the section entitled ‘Forwarding and Receiving’). To change the receiving port, disable the existing one and create a new one using the “new” button. Enter the new port and make sure to set the sourcetype to “process_history”.

Note that although this Utility generates a PIDataWriter.exe.config file for the PI to Splunk service, we do not recommend editing that file by hand; configuration of the service should only be done using the PI Data Writer Configuration Utility.

For now, all further documentation around using the Sideview Process Historian and troubleshooting common problems can be found within the pages of the app itself. We will revisit and expand these online docs soon with more content, screenshots and screencasts, but for now download it, set it up, and see for yourself!

What’s Coming In Future Versions?

Most importantly, the features that are coming are the features that people ask for. So if you havent sent us an email telling us what you’d like to see, do so now by emailing

However what we see as the next big feature to work on, is the ability to label time periods with categories of activity like ‘production’, or ‘cooldown’ or ‘test cycle’.

  1. Admin can add labels to slices of time — for example a particular 17 minute period could be labelled ‘production’, or ‘cooldown’ or ‘test cycle’.
  2. Admin can use a report builder interface to show any historical statistics and behavior by these labels.
  3. Admin can run a report showing current behavior as compared against recent baseline behavior for the given label.
  4. Admin can set up monitoring that lights up and sends emails when current behavior in a labelled time period deviates quantitatively from historical baseline behavior.
  5. Admin can set up monitoring to detect qualitative anomalies and not just quantitative deviations.