If you’re looking for backup of Office 365, you’re not alone. The recent surge of interest in Office 365 backup has come as companies realize that, for all great features included with Office 365, backup isn’t one of them. Whether it is SharePoint Online, Exchange Online, OneDrive, Teams, or the audit log, having a backup solution is critical to protect against accidental or malicious data loss.
This growing awareness and demand have given rise to a plethora of vendors all clamoring for a piece of the Office 365 backup market. This includes both small players and some vendors who are well established in the backup space.
But, the unfortunate reality is that most offerings fail to deliver on their marketing hype.
The landscape is littered with companies who purchased an Office 365 backup solution only to find that it was unable to work at scale, or failed to meet core requirements.
This post will examine some of the common pitfalls of Office 365 backup, with a focus on how HubStor’s Office 365 backup offering is unique at solving the challenges.
Problem #1: The agnostic problem
One of the most common pitfalls that vendors fall into is what I’ll call “the agnostic problem.” Established vendors have usually come of age and gained market share in a world viewed with tunnel vision. These vendors have focused for years on, and built complex platforms suited to a specific workload like VSS-based backup of on-premises Exchange Server, SQL Server, and SharePoint Server.
The truth is that Office 365 cannot be viewed myopically.
Not only are the Office 365 workloads very diverse, but how those workloads are accessed is highly specialized, with complex and unique APIs.
Indeed, Microsoft is trying to sell Microsoft Graph as the one-stop-shop for all data in Office 365, but they have a long, long way to go before they can make that vision even close to reality. The result is that no two data streams in Office 365 look the same, and some look EXTREMELY different (think SharePoint versus email).
Some Office 365 backup vendors have invested heavily to build platforms which, for example, are good at storing email, but struggle to support SharePoint and Teams data into that same platform. It’s a round-peg/square-hole problem, and it is crippling. Vendors in this scenario break out the duct tape and attempt to patch some solution just to make it work with their existing platform, but the results are never pretty. The resultant product exhibits all the classic deficiencies of a solution which was not designed from the ground-up to handle a variety of diverse workloads.
With all of the inherit complexities associated with attempting to retrofit new workloads into existing products, the only real solution is for the core platform to be built, from day one, to be workload agnostic. It sounds simple – indeed, you’d expect all software companies not to get hung up on this, but I can count on one hand, with fingers left over, the number of vendors I know of who have been able to accomplish this feat.
HubStor is one of those vendors, implementing a workload agnostic cloud data management platform from day one. The inherit flexibility baked into HubStor allows it to handle workloads from email to OneDrive, to VM backups, to application specific data pushed in from partner integrations or third parties, with no forcefitting required.
Problem #2: The scalability & performance problem
At HubStor, many of our customers rely on us for Office 365 backup, and many of these customers opted for HubStor after already attempting to use a product from an incumbent backup vendor.
Far and away, the number one reason we see customers adopt HubStor after attempting another Office 365 backup vendor is performance. These other products simply don’t work at scale.
Having spent my career working as a developer and architect in the Microsoft stack, I get it. Each data stream has unique APIs with a lot of domain expertise required to architect and implement a successful backup strategy.
Add to that all the nuances around how Office 365 throttles traffic, and the fact that you are working against a cloud system, and it’s not surprising we see so many vendors who are good backing up traditional workloads fall flat when it comes to backing up SaaS applications like Office 365.
If you have more than just 15 TB to backup from Office 365, then you need to pay attention. Just recently, a prospect told us that their Office 365 backup vendor was taking several weeks to process 2.5 terabytes out of a 13 TB scope.
One red flag is going to be the virtual appliance approach because these typically allow for one service account on just one VM.
Another red flag is the multi-tenant SaaS approach because it will only allocate a certain amount of horsepower to your backup, and when seeding the initial backup is when you need a dynamic model that can scale up/out the compute, as well as work intelligently with multiple service accounts.
The team at HubStor know Exchange, SharePoint, and OneDrive intimately, and have been able to translate that model into an Office 365 backup solution that works at scale, certainly in environments where our competitors have failed at just a dozen terabytes, but even at whale scale Office 365 tenants with a petabyte plus.
Take Exchange Online, for example. In production today, several customers see HubStor achieving sustained ingestion rates of 150-200+ messages per second, and we can scale things up even higher if faster performance is needed.
SharePoint Online and OneDrive present similar performance issues for most Office 365 backup vendors. The performance here typically measures in GB/day as opposed to item counts. A typical HubStor deployment can quickly ingest several TB per day from SharePoint Online and OneDrive, and we can scale much higher if need be.
Problem #3: The incompleteness problem
It might seem obvious, but it’s worth calling out another all too common limitation in the Office 365 backup space: Incompleteness.
By this, I mean that while almost all existing Office 365 backup solutions will backup certain parts of your Office 365 tenant, few provide complete coverage of all the essential data repositories in Office 365.
While most vendors have a story, at least in theory, around Exchange Online and SharePoint Online/OneDrive, few will mention anything about Teams, wikis, recoverable items, archive mailboxes, or the Office365 audit log. There’s a reason for that: Most vendors only capture active user mailbox data, and a limited set of SharePoint/OneDrive data, leaving administrators with gaps in their Office 365 backup protection.
Take the Office 365 audit log, for example. Backing up this audit log can be vital for some organizations, and yet few backup providers offer a solution. Without backup, your audit log data is permanently deleted after 90 days by default (unless you are paying Microsoft to extend this retention to one year). Of course, HubStor offers native support for backing up the Office 365 audit log as part of the solution.
As for SharePoint Online and OneDrive, while most backup vendors will claim support, the devil is in the details, and we see many vendors who can backup documents in document libraries or files in OneDrive, but nothing else. So, what about all your content in SharePoint that is not in a document library? In most cases, this data not in scope so you get no reliable recovery, putting you back at the mercy of the recycle bin.
In contrast, HubStor offers a broad backup solution for SharePoint Online and OneDrive, which can protect generic lists, system data, blog posts, wikis, and Teams data. This includes all previous versions of items in SharePoint, the permissions, and even things like attachments on generic list items.
Problem #4: The multi-geo problem
Recently Microsoft rolled out support for vastly improved geo-location control, which they have dubbed multi-geo. The introduction of multi-geo is a game changer for large multi-national organizations with complex data sovereignty requirements necessitating in-country storage of data. These multi-nationals can now finally move to Office 365, and stay in compliance.
Of course, for it to meet these same data sovereignty requirements, any Office 365 backup solution must also support multi-geo. The reality is that virtually none of the Office 365 backup vendors can provide such a multi-geo offering.
In many cases, the issue is that these Office 365 backup solutions are being delivered as multi-tenant offerings, meaning that data from various customers undergo processing and storage on shared resources. While this is less expensive for the vendor, it can make implementing multi-geo difficult because the vendor has pre-selected the physical locations in which their technology operates.
In other cases, the inability to deliver a real multi-geo backup is simply a result of the architectural complexity required to do so. Most systems are not designed to span multiple regions, nor include the ability to segregate data and data management seamlessly as needed. Virtual appliance-based approaches will claim to support it, but the administrative overhead is cumbersome at best.
HubStor takes a different approach that offers our customers the ability to scale a HubStor tenant across multiple geographies while maintaining a single pane of glass administration with the management of local admin privileges. Indeed, many of our large Fortune 100 customers selected HubStor because we have enterprise-level support for multi-region deployment.
And, of course, once you start talking about deployments across different regions, there often arises the need for regional administrators, i.e. admins who can manage resources and data in their region, but not in the other areas, or legal teams that have access to the federated data of all regions, or only specific areas as needed.
At HubStor, we have designed these kinds of region-specific controls directly into our authorization layer, making it easy to implement regional separation of concerns.
This solid foundation made it relatively easy for us to integrate tightly with Office 365 multi-geo.
Given that HubStor is already scraping Azure Active Directory for all users, the solution is simple and intuitive to apply, as highlighted in the steps that follow:
- Deploy resources into multiple Azure regions as needed.
- Configure HubStor’s Active Directory integration to harvest the multi-geo information (preferred data location) for users.
- Configure region specific backups that filter based on users marked as being in that region.
This solution ensures that data sovereignty is maintained automatically as users change locations, or new users are added or removed.
With HubStor, user data resides in the proper Azure region dynamically to maintaining alignment with Office 365 multi-geo settings.
Problem #5: The pricing problem
The vast majority of Office 365 backup vendors price their offering as a cost per user per month. While this is favorable to smaller organizations with a handful of accounts, it does not scale for large enterprises and educational institutions that have thousands or tens-of-thousands of users.
For example, let’s say you can get the vendor to reduce their pricing to $1/user/month, an Office 365 tenant with 30,000 mailboxes and OneDrive sites will be looking at $30,000 each month, or $360,000 annually.
For SharePoint Online, the per user pricing model is often not a great fit because there usually isn’t a large data volume to protect.
HubStor’s pricing model is a transparent consumption-based approach that meters off the underlying cloud pricing.
We don’t care about the number of users. Instead, pricing is based on the compute profile and the physical data volumes in the backup after deduplication and compression.
Using HubStor’s model, smaller businesses are usually not a good fit for their own dedicated HubStor tenant. But HubStor has a growing partner base with Managed Service Providers (MSPs), and they can help small companies use HubStor in a shared model.
However, medium and large Office 365 tenants do well in it, achieving economics as low as $0.17/user/month in some instances. To put that figure in perspective, that’s a reliable enterprise-grade backup at 2.13% the cost of an E1, or 0.85% the cost of an E3, or 0.49% the cost of an E5.
It is worth noting that HubStor’s pricing model has some variability. For instance, in the early months when seeding the backup, the compute profile typically scales up to expedite the protection process. Afterwards, things scale down for steady-state incrementals. So your cost per user economics, if you opt to view HubStor costs in that light, will improve over time.
Problem #6: The multi-domain problem
Here’s a common scenario we see with our Office 365 backup customers: Multiple operating entities, or through divestitures, mergers, and acquisitions, you need to backup multiple Office 365 tenants.
Does your backup solution allow you to sync with multiple Azure AD domains and connect to many Office 365 environments without headaches? How does their per-user pricing model work in scenarios like this? What are the administrative and licensing complexities?
With HubStor, we can scale out your dedicated HubStor instance to support a variety of scenarios that involve sub-domains, multiple domains, and different policies you might want to manage across different Office 365 tenants. And we do this without adding complexity with licensing or separate configurations that you administer.
Problem #7: The customer experience problem
You moved to Office 365 to get out of the weeds of managing the infrastructure behind Exchange Server, SharePoint Server, and other collaboration systems. But when it comes to Office 365 backup, some products will have you back in the weeds.
There are two common approaches, as follows:
The problem with a virtual appliance is that it doesn’t scale and you have to think about (and monitor and manage) the VM, it’s security configuration, updates, etc.
The problem with the multi-tenant SaaS approach for Office 365 backup, or any backup of a SaaS application, is that multi-tenant is too inflexible to solve for multi-region scale out, data residency, hyper scale data volumes, and performance.
In most cases, the experience of these options is going to be poor if you are a medium or large enterprise.
HubStor’s single-tenant architecture offers an alternative approach that gives you the best of both worlds.
On one hand, it is a fully-managed SaaS solution, so you don’t have any new software or IT infrastructure to manage. Secondly, it scales and offers the flexibility to meet enterprise-grade requirements.
Problem #8: The short RPO problem
Do you prioritize all data the same? Probably not.
In backing up SaaS data, especially with Office 365, we often hear that certain Teams and OneDrive sites are of particular importance. In other words, they contain critical data that is higher priority than other information.
Does your Office 365 backup allow you to assign different backup policies, or is it a one-size-fits-all approach?
For larger Office 365 tenants, this question is particularly important because even incremental backup jobs can take time. That means you may be pressed to fit incrementals in over a nightly run.
To solve this problem, HubStor introduced continuous backup which can capture changes as they occur in SharePoint, OneDrive, and Teams site collections.
Problem #9: The ‘can your recovery handle scenarios A, B, C, D, E, F, and G’ problem?
The challenge with enterprise backup is striking the right balance between simplicity and feature depth.
Backup vendors have traditionally delivered highly complex products. But now many of them are guilty of being too simplistic.
If the Office 365 backup solution is too simplistic, you will find yourself with limited recovery options.
In our experience, you will encounter a spectrum of recovery scenarios. Over time, listening to our customers, we have developed a flexible recovery engine that supports dozens of Office 365 restore scenarios.
Beyond Office 365 backup
The final category of deficiency we see with a lot of the Office 365 backup solutions is that they solve a single problem only.
Indeed, if all you are looking for is a backup and recovery for your Office 365 data, then maybe this isn’t a problem.
However, you may have data in other SaaS apps that you need to protect also.
And then there’s the backup of other workloads like VMs, on-premises shared storage, and PaaS cloud storage accounts like Azure Blob Storage and AWS S3.
Do you want to deal with a different vendor and separate licensing for protection of all this data?
Additionally, you can solve your archiving challenges on the same unified data platform, including immutable retention for compliance, file server cloud tiering, legal discovery, GDPR, application retirement, user portal access, and message journaling.
The bottom line: More and more customers are looking for a full-fledged data management platform, and not just a one-off backup. HubStor is emerging as the market-leading cloud data platform that converges backup and archive for workloads across the hybrid cloud in a flexible SaaS delivery that scales.
Key takeaways to help you evaluate Office 365 backup vendors
If you are evaluating the flooded space of Office 365 backup solution, then here are some summary thoughts to consider:
Does the product cover all the data you need to protect? Most vendors only capture active user mailbox data, and a limited set of SharePoint/OneDrive data, leaving administrators with gaps in their Office 365 backup protection.
- If you have a sizable corpus of data to backup, investigate performance claims because most offerings move too slow. Most offerings do not scale beyond a few terabytes of data to backup.
- What kind of recovery scenarios to you anticipate? Most offerings provide basic recovery options only.
- How is pricing as a percentage of what you are paying for the Office 365 licenses? Most vendors’ pricing will not scale for large user counts.
- Do you need to scale out across geographic locations, domains, and support multiple Office 365 tenants? Most vendors don’t support scaling like this.
- Are you forced into treating all of your Office 365 data the same when it comes to RPO and RTO? A good solution should offer the flexibility to assign priority backup policies to important data sets.
- What kind of experience do you want as a customer? Not all solutions offer a simple customer experience.