<?xml version="1.0" encoding="utf-8"?>
<rss xmlns:atom="https://www.w3.org/2005/Atom" xmlns:sy="https://purl.org/rss/1.0/modules/syndication/" xmlns:dc="https://purl.org/dc/elements/1.1/" xmlns:content="https://purl.org/rss/1.0/modules/content/" xmlns:wfw="https://wellformedweb.org/CommentAPI/" xmlns:slash="https://purl.org/rss/1.0/modules/slash/"
  version="2.0">
  <channel>
    <title>Skype for Business Server Real-Time Statistics Manager</title>
    <atom:link
      href="https://docs.microsoft.com/archive/blogs/skypestatsman/feed.xml"
      rel="self"
      type="application/rss+xml" />
    <link>https://docs.microsoft.com/archive/blogs/skypestatsman/feed.xml</link>
    <description>Tips and tricks for getting the most out of StatsMan</description>
    <lastBuildDate>Wed, 17 Apr 2019 03:36:00 GMT</lastBuildDate>
    <language>en-US</language>
    <sy:updatePeriod>hourly</sy:updatePeriod>
    <sy:updateFrequency>1</sy:updateFrequency>
    <item>
      <title>StatsMan Chart Drilldowns</title>
      <link>https://docs.microsoft.com/archive/blogs/skypestatsman/statsman-chart-drilldowns</link>
      <pubDate>Thu, 12 May 2016 15:11:18 GMT</pubDate>
      <dc:creator><![CDATA[Jason Balinski]]></dc:creator>
      <guid
        isPermaLink="false">https://blogs.technet.microsoft.com/skypestatsman/?p=75</guid>
      <description><![CDATA[As you probably already know, the root StatsMan page shows a great overview of your Skype for...]]></description>
      <content:encoded><![CDATA[As you probably already know, the root StatsMan page shows a great overview of your Skype for Business server environment usage and performance. Each chart on that page displays an aggregation of all the servers in your environment in real-time. So what do we do when we notice a spike or drop in those lines that is unexpected? How do we figure out which server is actually causing that anomaly?  That's where drill-downs come in.

<strong>Counter hierarchy</strong>

It's important to realize that there is an implied hierarchy of server populations in StatsMan, which is what allows much of the StatsMan interface to function as it does.  The hierarchy is laid out like this:
<ol>
	<li>* (Root population, all servers)</li>
	<li>Site</li>
	<li>Pool</li>
	<li>Server</li>
</ol>
If you're viewing a chart where all data is coming from the same server population, you'll be able to drill down to see all of the server populations in the next level below.  For example, if all series are from "*", you can drill down to all the sites, then drill down from a site to see all of the pools in that site, and then again to see the servers in one of the pools.

<strong>Drill baby drill</strong>

To the right side of any chart in StatsMan, there are a number of controls.  If all series in the chart are derived from the same server population, you'll see a drill-down option in that list of controls. The default view in StatsMan is a great place to start for this, since all of the charts will be using the "*" population.  Pick a chart that is interesting to you, and click the site drill down button.

You'll now be shown a page with the parent population at the top, and a chart for all of the sites below the parent. The scale of the child charts will sync together as they load to aid comparing their data.  You can also use the <em>compact view </em>button in the side bar to see more charts on the screen at the same time.

You'll notice that in the role view, the drill down button will change to <strong>pool</strong>.  You can now drill from any site to see all the pools within that site, and then again from pool to servers.  Note when you get to the server level, the population aggregation will be dropped since you're now looking at raw counter data from the server directly.

<strong>No limits</strong>

A great feature of StatsMan is that the drill down functionality is dynamic.  This means that you can drill down from any custom chart you create, as long as every series on that chart is from the same population.

&nbsp;

As always, if you have any questions about this or any StatsMan functionality, drop us a line in the comments or on Twitter @SkypeStatsMan.]]></content:encoded>
    </item>
    <item>
      <title>Getting the most out of the new scenario views</title>
      <link>https://docs.microsoft.com/archive/blogs/skypestatsman/new-scenario-views</link>
      <pubDate>Thu, 12 May 2016 14:57:45 GMT</pubDate>
      <dc:creator><![CDATA[Jason Balinski]]></dc:creator>
      <guid
        isPermaLink="false">https://blogs.technet.microsoft.com/skypestatsman/?p=65</guid>
      <description><![CDATA[What is the scenario view, and why do I want to use this?

With the release of StatsMan 1.1, the...]]></description>
      <content:encoded><![CDATA[<strong>What is the scenario view, and why do I want to use this?</strong>

With the release of StatsMan 1.1, the scenario view was introduced.  This is a view that is tailored to show you a huge amount of data in a small space, especially useful for certain <em>scenarios,</em> hence the name!

<a href="https://msdnshared.blob.core.windows.net/media/2016/05/Scenario-View.png"><img class="  alignright wp-image-105 size-large" src="https://msdnshared.blob.core.windows.net/media/2016/05/Scenario-View-1024x639.png" alt="Scenario View" width="1024" height="639" /></a>

<strong>How do I use it?</strong>

In the drop down menu near the top of the StatsMan homepage, you'll find the <strong>Scenario</strong> menu, preloaded with several scenarios we think you'll find useful.  Most of the scenario views are tailored to view a single pool in your infrastructure, with the exception of the <strong>Pool Failover</strong> scenario, where you'll look at a pair of pools side by side for easy comparison.

Choose one of the scenario views to get started. Note that the pool type will be pre-populated based on the scenario you're working with, allowing you to pick from pools that make the most sense for this scenario.  Choose a pool to view the statistics for it.  The statistics in the table will continue updating in real-time as long as the page is open.

Once the data is populated in the table, you'll notice that the chart area will remain blank.  You have a couple of options here.  You can click individual cells in the table to add data series to the chart to see their history, or you can click on the name of the counter to view the history of the average value for that counter for the entire pool.  You can add and remove as many counters as you wish.

<strong>Extra Credit</strong>
<ul>
	<li>To save space, values with too many digits in the table are compacted via rounding or scientific notation. Mouse over any cell to see the exact value.</li>
	<li>Click once to add a data series to the chart.  Click again to remove a data series.</li>
	<li>Shift-Click on a cell that's already added to the chart to highlight that data series on the chart.</li>
	<li>The control panel on the right side of the page still works on this page for any open charts - you can modify how much data history is shown on the page chart(s). By default, 30 data points will be shown.</li>
	<li>Even though a pool type is defaulted in the view, you can view the same statistics for other pool types, just choose a different one, and the list of pools will update to match your new chosen pool type filter.</li>
	<li>After adding some data to the chart, notice the chart link still exists at the top right of the chart.  Click it to open the chart in a new window with easy customization via the chart legend.</li>
</ul>
&nbsp;

As always, if you have any questions about this or any StatsMan functionality, drop us a line in the comments or on Twitter @SkypeStatsMan.]]></content:encoded>
    </item>
    <item>
      <title>Working with charts in StatsMan</title>
      <link>https://docs.microsoft.com/archive/blogs/skypestatsman/charts-in-statsman</link>
      <pubDate>Thu, 12 May 2016 14:26:37 GMT</pubDate>
      <dc:creator><![CDATA[Jason Balinski]]></dc:creator>
      <guid
        isPermaLink="false">https://blogs.technet.microsoft.com/skypestatsman/?p=85</guid>
      <description><![CDATA[There are a lot of features packed in to the StatsMan charting system.  Let's dig right in and...]]></description>
      <content:encoded><![CDATA[There are a lot of features packed in to the StatsMan charting system.  Let's dig right in and expose some of the best features.

<strong>Expand a chart to edit it</strong>

In the default and drill down views, you can expand a chart by clicking its title bar.  This will expose the chart's legend which will allow you to add or remove series from the chart.

<strong>Link back to any chart</strong>

The link at the top right of any given chart is always up to date, so after adding or removing series to a chart you can copy the link and bookmark it to see the same counters at any time.

<strong>Create a custom chart</strong>

The default view in StatsMan is great, but there are many, many more counters included than what is shown by default.  Create a custom chart by clicking the <strong>Chart </strong>menu, then click <strong>Custom Chart</strong>.  A new page will appear with a new chart with only one counter for CPU usage.  You can then customize the chart using the series controls in the legend below the chart.

<strong>Use a filterable list of all counters to add a new series</strong>

If you'd prefer to type (or paste) the name of a counter to add, simply click the "+" button to add a new line, then click the "Add..." button at the right side of the empty line.  A dialog will appear that will allow you to type to filter by all available counter names in the system.

<strong>Easily add a related series to the chart</strong>

Using the chart legend, it's possible to clone a series and modify only one part of that counter name to display a new, related chart series.  Simply click the colored button to the left of the counter name, and a duplicate row will appear below it.  Edit the portion of the counter that you want to change, and then click the Add button at the end of the row to add the new series to the page.

<strong>Highlight a series on the chart</strong>

Clicking the same button that clones the series line will also highlight a given series on the chart.  Clicking on a series in the chart will also highlight the same series.

<strong>Mouseover for details</strong>

Mousing over a data point on the chart will show you all available details for that data point, including the full counter name, local and UTC time stamps and the full decimal value for that sample.

<strong>Zoom/Pan the chart</strong>

It's possible to zoom/pan the chart directly to see details more clearly.  The zoom controls are located to the right of each chart.  Clicking the X button will allow you to zoom in and out on the X axis only, while the XY button will allow you to zoom both axes together.  To do so, click to enable the desired zoom mode, then scroll the mouse wheel over the chart.  Alternatively, clicking the background of the chart will enable X axis scrolling, and double clicking will cause the chart to zoom in. Deselecting either zoom option will restore the chart to the default zoom level.

<strong>Show data points on the chart </strong>

Setting the maximum points displayed in the control panel to &lt;50 in the default view, or &lt;200 in a view where there is only one chart shown will cause the individual data points to become visible on the plotted chart.  Zooming the chart window out to show the same number of points will also show points on the chart.

<strong>Download chart data for external use</strong>

While the charting system in StatsMan is extremely flexible, you may wish to download the data for use in another tool, such as Excel.  To the right of each chart is a button marked with a down arrow.  Click this button to download all of the data from the chart in an easy to use CSV format (hint, this is directly open-able by Excel).

&nbsp;

As always, if you have any questions about this or any StatsMan functionality, drop us a line in the comments or on Twitter @SkypeStatsMan.]]></content:encoded>
    </item>
    <item>
      <title>Store more data in StatsMan</title>
      <link>https://docs.microsoft.com/archive/blogs/skypestatsman/adjust-data-retention-in-statsman</link>
      <pubDate>Tue, 26 Apr 2016 12:19:05 GMT</pubDate>
      <dc:creator><![CDATA[Jason Balinski]]></dc:creator>
      <guid
        isPermaLink="false">https://blogs.technet.microsoft.com/skypestatsman/?p=45</guid>
      <description><![CDATA[So you've got a StatsMan instance running, and you think to yourself, "I have 64GB of memory but the...]]></description>
      <content:encoded><![CDATA[So you've got a StatsMan instance running, and you think to yourself, "I have 64GB of memory but the StatsMan Redis instance is only using 16GB of that.  Why can't I tell StatsMan to keep more data?"

It turns out you can. I will now give you some important warnings:
<blockquote>Directly editing the configuration files <strong>may cause StatsMan to malfunction</strong> in unexpected ways, give you unexpected data results or simply not start at all. If you choose to edit the configuration files manually, be aware that the settings you change <strong>may not be retained</strong> in the event of an upgrade. <strong>Always backup the configuration file(s)</strong> you are editing before beginning any modifications.</blockquote>
OK, now that we have that out of the way...

The default value for maximum sample storage is 1440, which just so happens to be the number of minutes in a day.  Increasing or decreasing this default value will cause StatsMan to <strong>scale near linearly</strong> compared to the initial value.  So, if you go from 1440 to 2880, you will now store two days worth of information instead of one, and use approximately double the memory.  The inverse will be true if you go from 1440 to 720, you'll store only 12 hours of data but use half the memory.

We would recommend you run with the default value of 1440 with your full agent load to determine how the Listener server behaves normally, and then <strong>tweak slowly</strong> to make sure your installation remains stable.

So, enough reading, how do we actually make the change?

First I'll plug my favorite text editor, Visual Studio Code.  Sure, you can edit the file in Notepad, but why not edit in style?  Using VS Code will color code the file for you to make it much easier to edit.  You can download VS Code <a title="Download Visual Studio Code" href="https://code.visualstudio.com/">here</a>.

Regardless of the editor, the file you need to edit is (default location):
<code>  C:\Program Files\Skype for Business Server StatsMan Listener\PerfAgentListener.exe.config</code>

Be sure to make a copy of this file before editing it.  Incorrectly editing the file will cause the service to fail to start.

Once you have the file open, search for the following configuration entry:
<code>  &lt;perfAgentListener redisService="localhost:6379" perfAgentListenerUriPrefixes="https://*:8443/"&gt;</code>

Add a property to this xml node called <strong>redisMaxStoreSamples</strong>, setting the value in quotes to your desired point storage count:
<code>  &lt;perfAgentListener redisService="localhost:6379" perfAgentListenerUriPrefixes="https://*:8443/" redisMaxStoreSamples="2880"&gt;</code>

Now <strong>save the file and restart</strong> the Skype for Business StatsMan Listener service.  Your new retention setting should take effect for each data series the next time a data point is saved for that series (typically within 1 minute of restarting the service).

We love feedback, let us know how it goes - we can be reached via the comments or on Twitter <a href="https://twitter.com/SkypeStatsMan" target="_blank">@SkypeStatsMan</a>.]]></content:encoded>
    </item>
  </channel>
</rss>