What a huge quarter we’ve had here at SentryOne! First we released Plan Explorer with some amazing new features and did away with the pay wall for PRO, and now we’ve finally released the next major version of our platform (v11), which has also been rebranded as SentryOne.
The purpose of this post is to cover the major changes to the platform, which definitely go deeper than the name change. We have new features to help you identify and analyze critical issues faster, we’ve revitalized the on-boarding experience for new users, we’ve combined features from different products to ensure you have everything you need for each platform you are monitoring, we’ve extended our integration with Tintri to provide both a deeper understanding of storage performance across the entire stack, and we've introduced a new SentryOne licensing option.
If you are upgrading a SQL Sentry deployment to SentryOne, you just hit the jackpot. Feature sets are now organized by platform support rather than by product SKU, and it means that you likely got a big upgrade!
I could write a long explanation of what that means, but I think a picture says it best:
All of our software tools were originally built to work seamlessly as a single platform, but over time, different features were carved out as separate products or add-ons. What we saw was a lot of people going without important parts that make up the whole. Now, you don’t have to decide which pieces you’re going to go without, because you get everything!
In the actual software, this translates to a more integrated feel, because there is no longer a need to split features between different products that target the same platform. You also don’t have to see menu options that refer to things you don’t have.
In purchasing, it’s far easier to determine how you want to cover your environment; you purchase the license that covers the platform you need to monitor. That’s the only decision you need to make. You don’t need to decide whether to leave the event management or the fragmentation management features behind, because everything is included.
For an upgrade, if your annual maintenance is up to date, you don’t need to make any additional purchases. Upgrade your license as you always have, download the new version as you always have, and run the installer to upgrade. Any relevant products or features that you did not have access to before will be available without any additional action on your part.
The bottom line is, if you are licensed for SQL Server, SSAS, Windows, APS, or anything else we support, then you will have every feature we have available for that platform.
Long ago we provided a Quick Start wizard to make sure certain things were in place for a brand new deployment. That served its purpose well, but it was created before we had extended features like Advisory Conditions and cloud synchronization. Raising awareness for those features was reason enough to rethink on-boarding, so you could take advantage of them right out of the gate, but we also wanted to create a more welcoming experience.
The new on-boarding wizard is more than a re-skinning of the Quick Start wizard. It was designed from the ground up to make sure a new SentryOne installation will have everything in place to let you take full advantage of the platform.
The wizard is intelligent, and only runs for the initial configuration, starting with a greeting from the new platform logo:
Creating the first contact user is much quicker, preloading details from your SentryOne license:
Configuring an email server is also straight forward, and it is now very clear that you can wait to do it later, once you have had a chance to plan your notification strategy:
I can’t show you a picture of this happening, but we’re also including our default Advisory Conditions pack as part of on-boarding now. We attempt to download the latest pack for you, but if we aren’t able to, you’ll get the latest condition pack that was available when the version you’re installing was released. All conditions are configured to go to alerting channels, which allows you to see them on performance dashboards and in the Environment Health Overview.
If you did choose to configure an outbound email server, we want to make sure you don’t get flooded with alerts, so we give you a chance to let us know what things you want to be notified of via email:
Last, but not least, you’ll have an opportunity to create an account at cloud.sentryone.com. This doesn’t configure synchronization for your servers, but it does give you a head start. You can also opt out of this, or any other configuration option, if you wish:
Once you are finished, you’re welcomed into the client for the first time, and prompted to add your first monitoring Target.
So, you’ve gone through the on-boarding wizard (or maybe you didn’t need to), and you’ve added a Target, maybe even several Targets. Now what?
Our intention has always been to provide a system that would help move people from a reactive to a proactive posture. Any time someone has to ask “what do I do now?” that process is slowed considerably. Environment Health Overview was created to make it very easy to determine what to do next.
To create the Environment Health Overview, we borrowed components from cloud.sentryone.com, and combined them with the deep performance analysis that can be performed using the full SentryOne client.
Quick navigation from the client start page has been replaced by a ribbon of icons. These perform the same functions they always did, like opening the global performance view, or quickly adding a Target or User.
A health score is provided, letting you know the overall health of your environment. In the screenshot above, you see that our test environment is at 72 out of 100. That’s not great, but it could be a lot worse.
The score is based on several factors, but put simply, the size of your environment and the severity, number, and age of Advisory Condition events contribute to the calculation. The age of the events is a key factor. Notice that in my screenshot, we know about over 1,200 medium severity events. If those had all happened moments ago, it would certainly have sent our score plummeting. There are, in fact, only a few events that happened very recently, and the health score understands that and takes it into account. While those older events exist, and may be worth reviewing, they are not affecting the health of your environment over the last 24 hours. The number of servers I'm monitoring also plays a part. Because I have a larger environment, I don't want a single server to send my overall score to 0, so each individual server is capped on how much it can contribute to the score. In a future update, you will actually be able to tell us which servers are more or less important to you, and we'll be able to weight those servers higher or lower depending on the weight you give to them.
You can use the time range drop-down in the ribbon area, shown at left, to change the current view and, by extension, the current health score calculation.
If I change to the last 15 minutes, for example, then suddenly things are looking considerably better for my environment.
The grid below lists a summary of events with the count, last occurrence, severity, tags, and, of course, the target that generated the event. From here you have some quick links that can be used to open the Performance Dashboard, Event Calendar, or the Advisory Conditions Events Log.
I can click or tap on one of the chart wedges, or on one of the severity and/or tag legends, to quickly filter the grid to those events or event categories.
Here I’ve used this quick filtering to get high severity events related to memory; this allows me to focus on two events in particular (out of nearly 200):
The data grid also supports custom filtering, sorting, column re-ording, and searching – just as you would expect.
This is only the beginning for the Environment Health Overview, and we welcome all of your feedback on how to make it even better.
The Disk Activity view has long been a favorite feature, providing for deep analysis of storage subsystem performance. We also added the ability to integrate with VMware and Hyper-V earlier this year, giving you the ability to see across the DBMS, OS and virtualization layers.
With this release, we wanted to tackle two more areas. First, we wanted to take our integration with storage vendor Tintri to the next level, revealing what is possible when we have direct API access to the performance metrics of the storage array itself. Second, we wanted to provide a historical visualization of storage performance across every layer, including the storage array.
If you’ve ever felt the pain of a storage bottleneck, and you found yourself arguing with 3 or 4 other people about who should be doing something about it, you can imagine how valuable this is. There is no more argument, you can point directly to the bottleneck.
Here’s an example:
This is from one of our Hyper-V servers, and I can see from view options at the top of the chart that I’m looking at the Tintri storage array.
I can derive several things from these charts, but the first is that the latency spike that came right before 11AM was not a problem for the storage array. During that time, the array wasn’t showing high latency. Incidentally, I would even be able to see if performance were being throttled on purpose by the QoS settings on the Tintri appliance. It registered in the hypervisor, but only half as bad as it did on the OS. Between the hypervisor and the OS, there was a bottleneck.
Next, I can see on both the IOPS and throughput charts that another VM on the same host actually caused the spikes. This spike in latency correlates with a period of high load from a specific VM, and this likely translated to a noticeable drop in IO performance across other VMs on the host that are sharing the same storage. A classic noisy neighbor.
Now I know exactly where to look, and who to ask for help if I need it. It took me longer to write about it than it did to figure it out.
These charts are incredibly useful even if you don’t use a Tintri storage array as well, and I’m really excited that we were able to get them into this release.
Watch for more visualizations like this integrated with features you are already using.
Finally, there is a great new option for Tintri storage arrays. You purchase a license that covers the storage array, and you are able to monitor any supported VMs that are stored on that array with SentryOne. This includes Win Sentry, V Sentry, SQL Sentry, and BI Sentry. Basically, the entire on-prem stack. To take advantage of this, talk to your Tintri representative.
Along with everything else we’ve done for the platform with this release, you’ll find that the integrated version of Plan Explorer is now up to date with the stand alone version. We have some big plans for this integrated version of Plan Explorer, so keep an eye on this area.
You may also notice several other improvements to help you get around in the client such as a reorganized navigation tree, and a redesigned global settings view. Keep an eye out for more big enhancements to the user experience in the future. We are going to be focusing on making all the features of the SentryOne platform much easier to find, and more intuitive to use.
The release of the SentryOne platform truly does mark a major milestone in our history, and we’ve put the benefit of our users front and center as we re-brand both our company and our product portfolio.
If you are new to the platform, I encourage you to take some time to have a look at the different areas I’ve outlined in this post. If you are upgrading or returning, you may have questions about how all of these changes affect your licensing or your environment. We will be in touch soon to help walk you through everything, but if you plan to upgrade before one of us reaches out to you, please feel free to contact us first.