Note- Dashboarding is completely subjective. Your target audience maybe a team of VCP’s or a NOC who does not know what co-stop is. Some may consider 90% a yellow, 95% orange, 98% red or 95% yellow and 97% red. Please use these as template for ideas and key metrics. (cherry pick what you think is the best for your use case)
The scope of this will cover how to create local dashboard account and a VM key metric dashboard using custom Views, Text and Scoreboards in vROps 6.4 (base 6.1 install, upgraded to 6.2, then 6.4)
The dashboard try to prove the application issue is “not on the VMware side” of the house and the app owner and OS team need to performance their own due diligence. The scoreboards motto is “green is good”. The chart below draws the same metrics from the past 24 hours to show any trends/spikes. Reclaimable capacity will help answer of the app owner demand 2 more vCPU’s when in fact we can reclaim 1 vCPU (pushing them to look for a runaway process/code change eating up the CPU)
If you know what you are doing; here are the links to import the Views, Text widget, Resource Kind XMLs and Dashboard.
https://vmscribble.com/x/VM-Key-Metrics-Dashboard-vmscribble.com-1-8-17.zip
Here is the end game dashboard (click to view full screen)
Create a local “dashboard” user account to author dashboards that will be shared out to AD users. Why? If a dashboard is created by an AD user and they leave the company, you will have to find their json files for their dashboards to import and re-share.
Administration / Access control / Roles / Green Plus sign
Create a new Role called Dashboard
Highlight the Dashboard Role and click the Pencil icon next to Permissions on the bottom right.
Check off:
Under Administration:
Resourse Kind Management (all 3)
Login Interactively
Under Content:
Alert Definition Management – (ALL)
Dashboard Management – (ALL)
Manage Metric Config – (ALL)
Super Metrics Management – (ALL)
Views Management – (ALL)
Manage Icon
View Content Actions Home Page
View Content Home Page
Under Environment:
Troubleshooting Management
View Dashboard/Env/Rec/Sum Home Pages
Click User Accounts / Green Plus sign
Click Objects / Role: dashboard / Assign this role to user checked / vSphere Hosts and Clusters checked
I am text block. Click edit button to change this text. Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.
Log out or open a different web browser or incogtnic mode to login with the local dashboard user account.
On the to right hand side, click on “dashboard” and user preferances. Change the Color Scheme to “Dark”
Create a title page by creating a css file.
http://www.w3schools.com/css/tryit.asp?filename=trycss_default is a great site to create and preview.
Content / Manage Metric Config / TxtWidgetContent / Green Plus sign / title / Paste in and Save.
Starting point since vROps is like trying to take sip out of an open fire hydrant. View all metrics for a VM and pick the ones you need. Click green icon so you are only displaying what is being collected. Draw the metric to see how it looks.
Detail / View. View some of the out of the box views to see how they were created and clone one to tweak if an out of the box one fits your needs.
As the local dashboard user, click Content / Views and the Green plus sign.
Name the View – VM CPU – 24 hours You can add the word “dashboard” to search for it.
Description – Workload/Contention/Ready/Co-stop %’s
Presentation – Trend. Max plot line of 25 is fine since we are not using it.
Subject – Search for Virtual Machine
Data- Search for workload. Double-click or drag CPU Workload % to the middle Data area. Metric Label- Workload %. Uncheck – Trend of the historical data / Forecast data for the next.
Repeat for the following metrics-
CPU|CPU Contention (%)
CPU|Ready (%)
CPU|Co-stop (%)
Click on Time Settings and change the date range to 1 day.
Visibility – Make sure “Dashboards through the View widget” is checked. The rest are subjective. I leave “Details tab in the environment” checked in case some is not using this dashboard and has the VM / Details/ Views screen open for the VM in question.
Select preview source will let you know if you picked the correct metrics.
Create another View for VM Memory following the same exact steps for CPU, but draw:
Memory|Workload (%)
Memory|Contention (%)
Memory|Latency (%)
Memory|Balloon (%)
////
Create another View for VM Disk Latency:
Storage|Read Latency (ms)
Storage|Write Latency (ms)
///
Create another View for VM Commands / Read+Write Rate
Virtual Disk|Aggregate of all instances|Commands Per Second
Virtual Disk|Read Rate (KBps)
Virtual Disk|Write Rate (KBps)
///
Create another View for VM Network Workload and dropped packets.
Network I/O|4000|Data Receive Rate|Workload (%)
Network I/O|4000|Data Transmit Rate|Workload (%)
Network I/O|4000|Received Packets Dropped
Network I/O|4000|Transmitted Packets Dropped
////
Create another View for VM Network Packet Rate:
Network I/O|Aggregate of all instances|Packets Received per second
Network I/O|Aggregate of all instances|Packets Transmitted per second
Scoreboard creation XML files.
Content / Manage Metric Config / ReskndMetric / Green plus Sign
Title- I used VM – CPU (replace CPU with MEM/DISK/NET) for the rest.
With the cursor on line 3 (between the <AdapterKinds>), click on <AdapterKinds> from the top menu and Choose the vCenter Adapter.
Click on Resource Kind and choose VirtualMachine
Click on Metric and select each one of these metric one at time (same ones from the View)
CPU|CPU Contention (%)
CPU|Ready (%)
CPU|Co-stop (%)
CPU|Provisioned vCPU(s) (Cores)
The next step is to fill in the Label, Unit and #s for the color. You can edit this on the screen or copy/paste into Notepad++, click language XML to make it easier on the eyes.
Know what #’s to fill in is subjective. Hopefully VMware will publish best proactive #’s as a starting point.
I added the VM vCPU count and set it at 8 for orange and 16 for red just to alert me this is a large VM.
Click on the Orange </> formate XML icon and click the Save button.
Repeat the steps for VM Memory. (open a new tab with the View open so you can see the raw metric name) I added the actually memory for the VM. (only measured in KB unless I create a super metric)
Repeat the steps for VM Disk. I added the # of datastores that makes up the VM for reference.
Repeat the steps for VM Network.
Create the dashboard. Content / Dashboards / Green Plus Sign
Title – VM Key Metrics
Seach for “text” and drag the Text Widget over. Click on the Pencil Icon to edit.
Title – Title
View mode – HTML
File – Browse – Pick Virtual Machine Key Metics and Save.
Drag the bottom right corner of the Text Widget to size make it fit the size of the text.
Search for the Reclaimable Capacity widget and drag to top right
Search for the Object List widget and drag below the text widget
Click on the pencil icon for the Object list widget.
Title – Search for your VM
Select which tags to filter – Expand Object Types and select Virtual Machine
Click on the arrows to expand Additional Column and click Pick Metrics
Adapter type – Virtual Adapater Object Type – Virtual Machine Metric Picker – Properties. Under Summary, double click Parent vCenter and Parent Host in the correct order (can not re-arrange) Fill in the Box Label.
The vCenter and ESXi host name are added to the far right column. You can click on an column name / arrow down and hide the other columns….but this is a per user setting in 6.3.
Drag 2 Scoreboard widgets over and make them half the of the screen vertically. Click the Pencil icon. The left will be used for the CPU and the right will be used for the Memory.
Title – VM CPU Current
Round Decimals- 0
Visual Theme – Pastel
Box Columns – 5 (the amount in the XML file)
Metric Configuration – VM – CPU
Repeat for the 2nd Scorecard for VM Memory.
Drag 2 View widgets and make them half the of the screen vertically. Click the Pencil icon.
Title – VM CPU – 24 hours – Click on each metric to hide/unhide
Views – Search for VM CPU – 24 hours and select.
Repeat for the memory view.
VM Memory – 24 hours – Click on each metric to hide/unhide
Drag a Scoreboard widget. Resize to to accoss the screen. Drag 2 View widgets and make them half the of the screen vertically.
Scoreboard – Used for the Disk XML file.
Round Decimals- 2
Box Columns – 6
Left view widget – VM Disk Latency- 24 hours.
Right view widget – VM Disk – Aggregate of all instances – Commands / Read+Write Rate 24 hours
Once again…just like the Disks, Drag a Scoreboard widget. Resize to to accoss the screen. Drag 2 View widgets and make them half the of the screen vertically.
This will be used for the network metrics. Round Decimals – 0
Left view widget – VM Network – Workload Dropped Packets – 24 hours
Right view widget – VM Network Packet Rate – 24 hours
Click on Widget Interactions +
“Search for your VM” will be used for ALL execpt for: Title & Search for your VM.
Click Apply Interactions and Save.
Click on the Gear Icon and Manage Tab groups.
Click the Green Plus icon. Create a new group.
Drag the VM Key Metrics Dashboard over to the new folder name.
Click on the Gear Icon and Share Dashboards.
Drag the dashboard name over to each group (one drag per group). You can check off “VC shared” if you want to really broadcast it…but it does come back to permission to view dashboards for those users.
Click on Home / Dashboard List and pick your dashboard. Search for a VM and click on it. All the metrics should automatically load. You can click on the arrow toward the middle of the left pane to give you more width.
