Problems with monitoring large Exchange and Office 365 environments
Whilst GSX Monitor and Gizmo can monitor a range of environments outside of just Exchange, this review naturally focuses on the product’s Exchange and Office 365 capabilities.
Now - I was being a little tongue in cheek above but some of it holds true. Granted, it’s easy to be certain about a few things. Firstly, if you currently monitor an on-premises environment using a tool that validates services are up and running on Exchange, or makes test connections to Office 365, it’s highly likely you can detect whether Exchange is available. Secondly, if you have a fairly simple set up – a single office or location – then you are likely to find performance, in general, will be similar for most users.
Increasingly IT administrator need a much more than that though to be able to manage their services effectively as the business now expects better reliability and has an expectation that should there be an issue, it can be pinpointed to the cause; and when there isn’t an issue, especially in the case of global firms, it’s a common ask for IT admins to want to know how to understand the ongoing latency from each region they need to support.
So, this is where GSX Monitor and Gizmo comes in very handy. They form components of an enterprise-scale monitoring suite that not only monitors Exchange or Office 365, but also monitor the third-party components like mail gateways and network devices, and crucially provide visibility from the client perspective using an innovative solution that performs synthetic transactions, like accessing mail or creating appointments, all from the client side.
The core use case for GSX Monitor and Gizmo are particularly suitable for larger organizations looking to gain visibility into a complex infrastructure.
Now, that’s a pretty sweeping statement but to elaborate, here’s a few of the common use cases related to Exchange and Exchange Online where I know a product like this is often asked for:
- Understanding client performance across multiple sites and regions. Either running on-premises or in the cloud (or a mix of the two) client performance is something you won’t typically have a lot of visibility of. Visibility into latency and transactional performance will be useful not only for establishing a baseline, but also when troubleshooting.
- Visibility into the health and performance of the end-to-end service. A move to the cloud means that there are many components that contribute to the end-user experience or can fail and cause problems with your service. Whether that’s AD, Exchange Hybrid, mail gateways, MDM solutions AD FS or Azure AD Connect – understanding the health of these as well as Office 365 is very useful to running a service with confidence and identifying the root cause of problems quickly.
- Performing synthetic transactions against Exchange or Exchange Online to verify service health. A highly available infrastructure can suffer failing components – but having a synthetic client send and receive mail, download data, create appointments, and perform other tasks that form part of the user experience are very useful in understanding whether the service remains healthy and performant.
- Collate data relevant to the service and allow it to be accessed by various roles. Collecting the data and performing tests is one thing, but it’s a common ask in larger enterprises for a solution that can make the information available to support teams responsible for managing the service. This is especially true in an organization looking to use a formal service management process and ensure that the relevant knowledge about issues within the service can be viewed by relevant staff, and where necessary alerts distributed to the right people.
These (along with a few other capabilities) are use cases that GSX claim the Gizmo and Monitor will help you accomplish. Over the course of this review – we’ll find out.
Setup and Installation
So, as usual the review is split into the basics of setup, install and configuration functionality and vendor support. However as this is an enterprise-focused product the setup is a little less relevant, so I’ll cover this a little more lightly than usual and focus a little more on how the products perform on the job.
Installation of GSX Monitor and Gizmo is straightforward. Both parts of the product require dedicated servers, the minimum specifications being provided by GSX within their detailed support and documentation site. Each installation requires basic pre-requisites, such as the .NET Frameworks, but no third-party or additionally licenced products like SQL Server are required in addition to the Windows licensing. All additional supporting components are automatically installed, and the resulting installation is ready for configuration.
GSX’s approach for our review was to provide first of all a product overview to help us understand the features and capabilities offered by the product, and then a series of remote sharing sessions to perform the installation and configuration along with an opportunity to understand how each feature works. This aligns to their service offering of a “kick start install” and appeared worthwhile.
This means that from a customer perspective you shouldn’t need to learn too much to get ready for an install – simply get the pre-requisites completed and the installation should be a breeze, whether that is performed in-house or with assistance from GSX.
Configuration and Functionality
For our review, I’ve built an Exchange Hybrid environment based on Exchange Server 2016, supported by Azure AD Connect and Active Directory Federated Services. To provide visibility into client access from different regions, I have clients located in two distinct regions; the UK and the USA. I am interested in gaining a complete picture of these services and ensuring that performance against Exchange Online meets user expectations when accessing the services from each region.
When configuring GSX Monitor and Gizmo it is necessary to perform tasks in several places; firstly, GSX Monitor must be configured with the source servers to monitor, along with the tests I’ll perform and important details like notification email distribution lists. Secondly, I’ll add in our regional client robots into the GSX Monitor server. It will deploy remote tools to those clients to communicate with and orchestrate synthetic tests. Finally, the GSX Monitor server must be registered with GSX Gizmo, which will provide access to our regional teams so that they can access the web interface as necessary.
After installation, our first task is to add the Exchange Hybrid organization into the GSX Monitor console. To perform this action, add a new organization and enter credentials to connect to the relevant Exchange Servers. To facilitate this, I’ve pre-created a service account with organization management permissions on-premises and global administrator privileges within the Office 365 tenant.
After addition of the Exchange organization, the Server Settings tab allows us to explore, and add, additional servers and services to ensure the full view of our organization is present. In addition to Exchange Servers, SharePoint on-premises and online, Skype for Business on-premises and online, Active Directory, ADFS, Azure AD Connect and other services can be added. The list is relatively exhaustive, and includes third party services and servers including Blackberry and IronPort – the type of add-on products typically seen in larger enterprises.
As part of the process to add each server or service, additional configuration options are present. For most types of server that can be monitored, suggested Windows Services are available for monitoring. In the case of Exchange Server, interfaces into Managed Availability are automatically configured to ensure that relevant information reaches the system.
Within the Server Settings tab, one area of note is the area we can configure our robot clients to perform synthetic transactions in various regions. This is known within the interface as the Exchange User Experience Manager. In the example below I’ve created a few core pieces that enable us to monitor. Firstly, I’ve created a configuration for the test, which in the example below is shown as O365. This defines the synthetic tests to perform that are intended to be representative of what a normal user will do, like download an attachment or create a new email. I’ve then created robot users, which within the interface is the name of the client (such as UK Central or US West to ensure we have a friendly name in the interface) and connection details so that GSX Monitor can connect to the remote system, install the necessary components to perform synthetic transactions and then the orchestrate those changes:
The Scenarios tab provides synthetic based tests from the central server to verify functionality. These tests include for Exchange Server, relevant tests for mail routing, mailbox access and mobile device connectivity.
The mail routing scenario will be useful to many organizations as it allows us to configure many mail flow tests to ensure that messages route both within the organization and externally. In the example below, I’ve configured several tests to verify mail flows between Exchange on-premises and Exchange online. Depending on the environment SMTP tests are available in addition to the capability to use Exchange Web Services to send or test receipt of messages. It is easy to envisage this test being configured in many organizations to verify end-to-end mail flow through external or third party cloud gateways, between Exchange sites or other internal email systems as well as Exchange Online.
The Exchange mailbox scenario provides a deeper view into the capabilities a user has outside of simply receiving an email into their mailbox. In the example below, I’ve created a test to connect to Exchange Online, connect to a mailbox, create a new folder, and then create a new task and create a new email. The tests available are very flexible, even including the ability to add attachments to the test messages. As with the mail flow scenario, several tests with customizable actions can be created. The ActiveSync tests provide similar functionality.
In addition to configuring monitoring of various services, reporting options can be configured. This is known as the Spot Check Report and allows configuration of regular scans of mailboxes, public folders, and mobile devices in both Exchange Online and Exchange on-premises. These reports are useful to track where mailboxes are hosted, quota limits, last connection times and other useful data that would normally be collected manually using the Office 365 portal or customer PowerShell scripts. The data collected can then be output in different report formats. The example below shows a snippet of the full report:
Once the configuration for GSX Monitor is complete and the system has had an opportunity to collect data, there are several methods available within the Monitor interface to view data. Note that this information is also available in GSX Gizmo – we’ll check that out in a moment though.
Our primary point of call is the Main View, which is the start-up page when launching the GSX Monitor console. This allows us to organize the relevant resources and get a red/yellow/green view of the overall infrastructure for rapid pinpointing of services that may have issues. In the example below, we see immediately that one of our AD FS servers has an issue that needs further investigating, and a warning is shown against the health of Office 365 and from the Exchange User Experience manager.
The Statistics tab allows us to drill down into details about Exchange Servers on-premises. In the example below we can view the health of the server itself, services running, protocols available alongside other key information such as network traffic, client response time as reported by the server, SMTP messages, mail queues and database information.
The Graphs view allows us to look at the historic performance and availability data collected from services and view this over a time. In the example below one of the more basic graphs is shown, providing historical data about Exchange Online access time.
Also of note is the Alert viewer within the GSX Monitor interface. These alerts are likely to, depending on your requirements, be received by email, or pushed to your helpdesk system. The Alert viewer itself allows viewing of the historical alerts the system sends based on the notification and alerts chosen for individual services or at a higher level, such as in the default policy. The example below shows a range of alerts including health alerts collected from the Office 365 portal, user experience issues and issues outside of Exchange Online or Exchange on-premises such as AD FS server issues where enough RAM is not available.
To provide access to those who don’t have access to the console of GSX Monitor, it is necessary to configure GSX Gizmo.
The first thing to notice about GSX Gizmo is it’s simple to use. Upon accessing for the first time, a prompt to add a source is displayed. What isn’t very clear is what that means – thankfully GSX support clarified very quickly that this should be our GSX Monitor servers.
After adding the GSX Monitor server to GSX Gizmo, all the configuration completed prior to this is available in GSX Gizmo, along with the historical and future data and statistics.
This then makes it very easy to first organize the views in GSX Gizmo to provide relevant data to relevant teams; for example, to create views by location or platform to mirror the way you support the services themselves. Then support engineers can access the system and view alerts and relevant performance information.
When we scroll down and view data available we’re offered quite detailed graphs providing relevant information, along with exact statistics available by hovering the cursor over the timeline. In the example below, from the same screen showing robot user experience data, we can see the time taken to perform various tasks in Exchange Online. What’s obvious on the graph is at some point recently the time taken to download an attachment took a lot longer than usual, although most other tasks performed well.
Pricing and Support
Pricing for GSX – like many products aimed at enterprise customers is likely best discussed with directly as the exact bits you will need will most certainly affect the price. However, pricing on GSX’s website at list price is very competitive when compared against competing solutions that don’t offer as much functionality. Additional services from GSX to hold your hand during setup and configuration will cost extra, but from our experience will be worthwhile.
Support during the review was excellent and GSX went to great lengths to ensure they were available with the right resource to provide assistance, including expert support staff that could not only help us configure the service but also provide important and useful guidance on the best way to configure the services for our environment. Of particular note is the support documentation available. For most the technical set up tasks, like preparing robot machines for monitoring, documentation was easy to follow.
It’s hard to find fault with GSX Monitor. It meets the types of requirements that are often asked about by IT Pros looking to get visibility into Office 365 or their Exchange Server environment, especially in a global environment with many interconnected systems.
The killer feature that is worthy of attention is the robot user feature which was far simpler to set up than I could have imagined and worked very well.
The GSX Gizmo interface makes a massive difference to a product like this and will make it very easy for global IT organizations to adopt the products and streamline the Exchange and Office 365 support process.
Overall, it’s a great product and one I expect I’ll be certain to recommend on a regular basis.