This post is in response to an InfoStor article entitled Cloud Storage Pros and Cons.
I read a Nasuni blog post yesterday that’s a response to a recent InfoStor article on the pros and cons of cloud storage. The blogger at Nasuni gave a nice rebuttal to some of the cons. I’d like to provide HubStor’s response as well. The bottom line is that the cons listed in the article, while true for many cloud storage approaches, are not at all true when your cloud storage strategy uses an enterprise-grade solution like Nasuni or HubStor. So let’s take a look at how HubStor helps IT teams overcome InfoStor’s cloud storage cons:
Con – Lifetime Costs
The InfoStor author says that “your costs over time might increase [in the cloud]”, but there’s no quantification to support the claim. In our analysis, cloud storage is by far the more cost efficient long-term storage strategy for enterprises.
The truth is, after you’ve felt the pain of the big capital expense to acquire traditional storage, the in-house storage cost model does look great in the next few following years, but then the hardware ages and you’re hit with a hardware refresh and data migration – and your cost model starts over with another spike in capital expense. Vendors like Pure Storage and their Evergreen model are making this less painful, but in-house storage still presents many other cost factors that you simply don’t see in the cloud (backup, power & cooling, support contracts, etc.).
Nasuni claims their customers see 40-65% cost savings in the cloud relative to traditional storage. With HubStor, we claim around 66-75% cost savings. If we configure your HubStor tenant with Azure Cool Storage, it’s even better cost savings. This is by no means a knock at Nasuni. An ultra-high-end HubStor configuration could make us more expensive than Nasuni (but still showing cost savings compared to traditional storage long term).
We’ve blogged about the misconceptions around long-term cloud storage costs a few times, taking on cloud’s data transfer costs, comparing the fully-burdened storage cost models, and looking at the true costs of in-house storage. I think the “lifetime costs” argument is baseless. Other than our Financial Business Case for Cloud Archiving, I’ve yet to see someone really lay out the numbers.
Cons – Security
On the topic of security, some valid concerns are raised. Indeed, with many cloud storage approaches, security is not where it should be, and so it’s a hurdle for many enterprises that wish to adopt cloud storage. But at HubStor, we’ve invested heavily to achieve market-leading, enterprise-grade cloud storage security. Here’s how it stacks up to each security concern raised by InfoStor:
- “managing permissions (both internal and external)” – Many cloud storage approaches do not inherit or synchronize the source permissions, nor do they make it easy to manage access in the cloud. HubStor makes it dead simple. We do a complete capture and ongoing synchronization of internal ACLs on everything you archive to the cloud. This includes user and group permissions (with full group expansions), deny permissions, and permission inheritance, at both the folder and item level. In the cloud, we give you the power to run searches or policies by data owner, user access, and group access, making it easy to interrogate access rights. Through our RBAC feature, if you authorize select users with HubStor’s file sharing capabilities in the cloud, there is visibility and auditing for all internal and external sharing. Shares can be revoked and set to automatically expire, and with our built-in DLP you can block certain data from being shared. We expect to enhance access governance capabilities in the future. Even today, managing permissions in the cloud with HubStor is seamless and simple -- and in my opinion, much easier than traditional storage.
- “vendors accessing your files for marketing and encryption purposes” – With HubStor, we make it clear in our Master Subscription Agreement that we do not access your files for marketing purposes. Statistical data is gleaned for billing and performance monitoring. Microsoft also states in their documentation that they do not access your data. While this is a concern with many cloud services in the consumer space, it is not with true enterprise-grade archiving services such as HubStor on Microsoft Azure.
- “confidentiality of intellectual property” – HubStor is data-aware with its own built-in analytics, search-as-a-service, and sensitive data detection. If you happen to archive some sensitive IP, you can identify it with HubStor and lock it down with our DLP capabilities, and easily get it out if you don’t want it archived in the cloud.
- “transfers when on WiFi” – I won’t spend much time on this, other than to say I never use public WiFi, and security issues around WiFi are hardly specific to cloud storage.
- “synchronization (which document is accurate and are any incorrectly overwritten?)” – HubStor performs a hash-based verification of items as part of the ingestion process to the cloud. In the cloud, HubStor is a read-only archive with the option to enable versioning. No synchronization issues here.
- “multiple public cloud providers” – Sure, it’s more of a headache if you’re managing your infrastructure across multiple cloud providers. But the cloud storage con is not clear.
- “accidental deletion on shared drives” – This isn’t a cloud storage issue, but in the case you did accidentally delete a shared drive, HubStor will bail you out. Your HubStor cloud archive includes the ability to recover any data that you’ve archived (provided the administrator has authorization to use the export/recovery utility).
- “and lack of clear-cut audit trails” – This is definitely a problem with most cloud storage approaches, but not with HubStor. Everything is always audited in HubStor. All administrator and end-user activities are audited. System actions too. We make it easy to dig into the audit data and generate reports. You can also zero in on a particular item and see all of the policies that apply to it, its access rights, and all of its retrieval and sharing activity.
Con – Speed
The InfoStor author calls out download speeds as an issue. It’s true, cloud storage has more latency, so it's not a good idea for active data and applications that need low latency access to files.
With HubStor, you have cloud-integrated, policy-based archiving capabilities against your existing storage to decide what moves to the cloud archive, what is stubbed, and what exists solely in the archive. Customers typically use LastAccessedDate in their policies – stubbing content only when last accessed is older than 30 days.
When data is only located in the cloud archive and you need low latency access, simply retrieve it, save it locally, and use it from there. Eventually the item will find its way back to the archive when it becomes inactive again.
Con – Compliance
We agree with the InfoStor author that compliance is something to pay close attention to. With HubStor, businesses with compliance requirements can use public cloud storage and still meet their compliance needs.
Compliance requirements vary, but in general, HubStor’s features around auditing, immutability, access control, PII detection, and support for multi-factor authentication make it so that you can easily meet compliance while archiving securely on Azure public cloud storage. For example, HubStor provides actual WORM immutability over Microsoft Azure, meeting many compliance requirements around records preservation, such as SEC/FINRA Rule 17a-4.
Then there’s the issue of data sovereignty. HubStor works exclusively on Azure, and Microsoft is rapidly expanding its list of Azure datacenters around the world. When we deploy your dedicated tenant, we have the flexibility of deploying you into the Azure region of your choice. And for large multi-nationals, HubStor also supports scale-out across multiple Azure regions, allowing you to meet the data sovereignty needs for different countries under one HubStor configuration.
Con – Noisy Neighbors
The InfoStor article calls out noisy neighbors as a cloud storage problem. If your solution is deployed to a multi-tenancy architecture, which I believe is the case for Nasuni, then you’ll have potential for this. But there are other approaches in the cloud that avoid this problem.
At HubStor, you’ll never have the noisy neighbors problem because we always deploy each customer into their own isolated tenant. And that’s true whether you subscribe under HubStor or we deploy your HubStor subscription into your own Azure account. Our approach is a little more expensive because you’re paying for dedicated infrastructure in the cloud (beyond just storage), but it provides consistent performance and ensures no comingling of your data with other clients. The single tenancy model is well worth the extra cost in our opinion.
I think some of the cons called out in the InfoStor article are reasonable because they apply to most cloud storage approaches. However, technology is fast advancing in the cloud, and new solutions like HubStor are making enterprise information archiving on public cloud storage suitable for companies with strict security, eDiscovery, and compliance needs.