Monthly Archives: December 2014

vRealize Operations 6: Top-N widget

The Top-N widget is used extensively in vRealize Operations 6. The default dashboards use them. For example, the Host Overview dashboard use it, as shown below:

Top N is average 0

Notice in the above screenshot (click if not clear) that there is a little text 24h. That means 24 hours. You can adjust, albeit manually, the time period. The value you are seeing in the above Top-N is the average of an entire 24 hour period. So if there is a peak during the period, it may get flatten. It is also the last 24 hours, not yesterday or today. Checking at 9 am or 6 pm will give you a different result. If you check them at 9 am, you’re looking at 9 am yesterday until 9 am today. You are not looking at yesterday (0000 – 2400).

Because the Top-N is an average, you may want to know a bit more details. This is where the Sparkline widget comes in. Clicking on any of the Top-N will show the corresponding object in the Sparkline widget.

You can certainly change the value from last 24 hours to any time period that fits your business needs. Changing the default dashboards do not impact the way vRealize Operations works (e.g. its dynamic threashold calculation). Dashboards are just way to present information.

You can also create your own dashboard and do your own style. For example, I do not use Sparkline as I like to have greater detail. I use Line Chart. I also use Line Chart first, then Top-N second. So it’s the other way around. This is because I have a preference to see details. I use Top-N when I need to zoom into a specific time line, to reveal the objects giving me the value in the Line Chart. I use this Line Chart + Top-N combo a lot when working with customers. You will find many examples in my book.

If you are curious if the Top-N value is really average of the selected time period, you can easily test it. I created a manual group. It has only 2 members. They are the 2 VMs shown below.

Top N is average

There are 2 Line Chart widgets in the above screenshot. Each of them has a corresponding Top-N widget below it. The first line chart shows a longer time horizon. I chose 7 days. From here I could see that there are some spikes. The Top-N, however, does not reveal that. This is because it is an average. It has flatten the data.

The second line chart shows a much shorter time. I have zoomed into 29 Dec around 1 pm. The line chart shows that BCDR-Prod-SRM-Server had a spike around 50% and then dropped to 4%.

Since I know the time period, I’m going to configure my Top-N to zoom into that specific time. You can see below that I’ve configured to 12:55 pm – 1:05 pm. So I’m taking only 2 values.  Top N is average - 2

Since the first value is 50.17%, and the second value is 4%, we would expect to see a Top-N showing 27.085%. And you got it right, the Top-N shows 27.085%.


Top N is average - 3

Looking back. Looking forward

Looking back, 2014 has been a good year for IT professionals earning a living with virtualization as their main skills. Virtualization and cloud computing continue to grow. While there are still a huge amount of physical servers installed base, x86 based VMs continue taking share in the datacenter. The physical servers in this context includes both x86 and non x86 (UNIX and Mainframe).

Non-x86 still accounts for around US$ 8 billions a year in 2014, so as virtualization professionals we still have a lot of target to work on. It will take many good years. The migration to X86 are certainly more complex than standard P2V, as application migration is required. But that’s also makes it more suitable for seasoned and experienced IT Pros. If it was easy, your job would have been replaced by a much more junior guy. There is already a post that drives that message, so I’m not going to elaborate.

2014 saw Virtualization entered the realm of Storage and Network. Yes, I know it certainly happened before 2014. In 2014 it became much clearer that Storage and Network will follow Server. They will all be virtualized. I think it won’t take Storage and Network very long, as the vendors & industry have learned from Server virtualization. This is a good news for us Virtualization Pros, as that means the job scope is widening and the game becomes more interesting.

Looking forward, I think in 2015, it does not take a prediction to see that NSX, VSAN, EVO will grow fast. It’s now a matter of projection. At my personal level, I’ve had customers buying NSX and VSAN in 2014. From my discussions with customers and partners, more will buy in 2015 as the technology becomes more common. The upcoming release will also help. I like what I see in the 2015 roadmap.

Related to virtualization, I see that Fibre Channel will continue its decline. 10 GE and SSD make the case for distributed storage compelling, especially at the entry level and midrange. Enhancements in vSphere in 2015 will also strengthen the case for IP Storage.

There are also other areas that virtualization can address better. One of them is Management. As datacenter becomes to have more VMs than physical servers, the management tools have to be built for virtualization. It has to know the differences between VM and physical servers well. In 2014, I’ve worked with quite a number of customers to help them operationalize vCenter Operations 5.x. Once the dashboards are tailored for each role, they found v5 useful. vRealize Operations 6 brings many enhancements that customers will find the product even more indispensable. I like what I see in 6.0. Knowing the roadmap for 2015, I think it’s going to be better (more complete features, easier to use, etc).

All in all, I am excited about 2015. The evolution continues, and I think the pace is accelerating. What started as server virtualization many years ago is now touching the entire datacenter.

What’s your take for 2015? How do you think it will impact your career? How do you plan to take advantage of the expansion of virtualization to Storage, Network, DR, Management, etc.? I’m keen to hear your thought. You can drop me an email at e1 at vmware dot com.

Have a great adventure in the bigger Virtual World in 2015!

[Update: I did a follow up post here]

vRealize Operations 6 redirect to syslog server

The steps to redirect to an external syslog server (e.g. Log Insight) has changed in version 6. You can do it via the UI, and there are 2 places to set.

The first one is for the main log. Click on the screenshot below to see where the log is configured. I only have 1 instance of the vRealize Operations (which I think is enough for most use case). There are different components of vRealize Operations, and each has its own log configuration.

20 syslog

You need to click on each of the component one by one. You cannot set it at the instance level. The “Settings” icon is greyed out when you select the instance. In the screenshot below, I selected the “Collector” component. I then clicked on the “Settings” icon. It opened a dialog box, as shown below. From the dialog box, you just need to specify the syslog server. Port 514 is the default port for syslog.

21 syslog - gotta do one by one

You need to do it for every component that you want to log. In the screenshot below, I did it for the “web” component.

22 syslog - gotta do one by one

Note that not all components have log. You will get a pop up telling you that there is no log option, as per the screenshot below.

23 syslog - most do not have

All the above is for the product logs. You should also track the security logs, which are configured in a different place. Follow the screenshot below. Click on the icon (I highlighted in red). A small dialog box will pop up, as shown below.

23 syslog - Audit has a different UI

That’s it!

I use Log Insight as my syslog server. You can see that the data showing up in the screenshot below

24 syslog - result - need management pack

At the time of writing, there is no content pack yet for vRealize Operations 6. The content pack covers version 5.x