Design Inputs needed for vC Ops Custom Dashboard

[vkaranam]

Hello Guys,

It's almost Holiday time and happy thanks giving to all. These communities are helping a lot and the reponse for the issues was awesome.

I have a Scenario as explained below and guys i want some inputs for the design.

I have around 4000 vm's,  200 Hosts, 1000 Datastores. I want to design the custom dashboards for different teams like Citrix,vmware,NOC & Storage. I know how to use most of the dashboards but my concern is i need some design inputs as it is a huge environment. What are main aspects that everybody are looking for building the dashboards to those teams.

Do we need any external adapters which might help getting some extra value to the dashboards?

Thanks

VK

[vcpguy]

There are multiple things that we need to look here. I have seen similar environments and everybody's request and needs are different. Having external adapters always help. But, you should be able to get value out of it. To start with, you can go with

1. Dashboard for Noc
2. VMs with large disk space that are not using their allocated disk space and you can reclaim disk space back
3. Host uptime for the entire environment
4. Top 100 vms
5. Top 100 alerts
6. Ldap access for all the users that are going to access vcops
7. Data store io information
8. Vms having io issues
9. Data store having io issues
10. CPU and memory contention
11. Dashboard for application
12. Manager dashboard, which shows Heath of key vms or entire environment
13. If, you can associate vcops with remedy or any other ticketing system, that will be great too.

Hope this helps

[admin]

Hi  vKaranam,

My blog is www.virtualclouds.co.za. I have a few ideas there for you...go through all the pages, there is a lot with examples.

[MattG]

I would break down between Operational and Application dashboards.

• Operational -
  • Contain vCenter adapter metrics configured in dashboard based on how they operate.
  • Examples:
    • Disk dashboard that contains only disk/datastore metrics and alerts.  Could use Datastore tags to differentiate datastore based on what is meaningful to them.
    • Operational Overiew by region.  Would contain all interesting operational metrics in a high level dashboard
    • The Generic Scorecard is a useful widget to map to customer/app/operational important metrics.
• Application
  • Creating applications using only the vCenter Adapter is better then what they have now,  in that it will group the resources together in a more meaningful way.  However,  to get a real understanding of applications, monitoring tools and adapters like SCOM or Hyperic are important as they will provide OS and application metrics which complete the picture.   I would imagine that is why Hyperic is being included with certain versions of vCops 5.6
  • Organizations usually do not put alot of thought into monitoring applications as groups of resources,  their relationships, and monitoring the whole stack including app perf data.   You will find that this conversation will force them to ask and answer questions they hadn't thought of before because there was no good solution to monitor all levels in one tool.  vCops change that.

-MattG

[bryansemple]

I am out till Monday morning for the Thanksgiving Holidays.

[vkaranam]

Thanks a lot guys for your inputs. I will try these and let u all know how it's going.

Thanks

VK

[vkaranam]

Hey Hugo,

I am looking at your website and i found this

vC Ops : Dashboard with Costs per Cluster

Can you explain with waht widgets you have desgned this dashboard and what kind of super metrics have you created.

Thanks

VK

[admin]

Hi,

Widgets is the Scoreboard.

Super metric is : (Capacity Remaining -> Count -> VM * "your costs) )

((min($This:A1434)) * 500) + ((min($This:A1553)) * 500)

Copy/past this formula into the super metric editor and you will see the metric's I have used.

Let me know if you need more help !

[vkaranam]

Hey Hugo,

Do we need to create a supermetric package in the dashboard to apply this formula?

What is This resource in the Manage super Metrics.

Can you explain the formula in depth?

Thanks

VK

[MattG]

V,

You always need a Super Metric Package.   Super Metrics get added to SM Packages which get added to resources.

-MattG

[vkaranam]

Hello Matt/Hugo

I have created the super metrics and applied to the packages and also i applied it to the generic score board. While creating the super metrics i have used the Host Usage, Read , write in the disk. i applied sum and everything is fine. when i am applying in generic score board it is showing 2701.733 kbps in red color. how can we set the read /write and usage values of the disk in the generic scoreboard.  I mean the threshold values (green,yellow,orange and red values) as marked red in the image. What are the default kbps levels?

Thanks

VK

[MattG]

V,

I am not exactly sure what you are asking?   You obviously know how to set the ranges in the Scoreboard based on the first example.

-MattG

[vkaranam]

I am asking how to setup the ranges(green/yellow/orange and red) if we are measuring the read/write and usage of the disk in kbps.

if you see my image the capacity|worklaod is measured in % and the Supermetric|Sum of Host R/W/U is measured in kbps. What are the values that needs to be entered (syntax) for the Green,yellow,orange and red levels.

Thanks

VK

[MattG]

You are trying to match the Disk IO Workload % so that you can determine what 100% equals in KBps?   If so,  you will not get them to match at the disk workload is a dynamic number that is redefined everytime a new high water mark is established for the Disk IO.

I guess if your vCops is running long enough the Disk IO HWM will stabilize,  but until it does that % is useless as it is constantly redefined.

If you wait for the DT to start appearing for the new SM,  it eventually will give you an idea of the max and min #s for the metric.

-MattG

[vkaranam]

I am not trying to match the workload% and the Spermetric i used. In the image there are two different metrics Capacity|workload and Supermetric|Host I/O R/W/U . For the Capacity | workload  i used the % ranges 40, 40-60, 60-80 and 80 for the green,yellow,orange and red levels. In the similar way which ranges i have to use when i set the KBPS as the measured unit for Supermetric| Host I/O R/W/U .As of now i am keeping some dummy values(like 40, 40-60) for the Green, yellow, orange and red ranges even they they are measured in KBPS so that the generic scorebaord can consider collecting inofmation for the Supermetric|Host I/O R/W/U . From your point  DT will take care of all the KBPS rates for the disk and it will set its own ranges, which is fine. How will it(DT) alert the user regarding the high or low disk rates without setting up the green, yellow, orange, and red ranges.

Thanks

VK

[MattG]

The Scoreboard doesn't alert,  just updates the colors on the screen.  You would need to determine what the #s are for the colors based on your usage patterns.  To have hard thresholds applied you need to go into the SMP and click on the SM and then Advanced.  To have the Advanced setting send email alerts you need to modify an XML file on one of the VMs.

-MattG

[vkaranam]

ya i got  what you are saying. It makes the connection.I know about that XML file modification.But it's not that easy to set these advanced levels and xml configuration. do u know how to modify the xml file(what changes in the file will generate the alerts)

Thanks for your info Matt.

Thanks

VK

[admin]

What you are asking is around how do vcops send an alert when the "red value" is reached ?

[vkaranam]

Ya Hugo. Also we had this conversation before for modifying the  XML file in previous posts. I need some guidance for modifying these XML file to send the alerts bcoz i am using super metrics and we have to include our own alerts if we use the Super metrics.

Thanks

VK