Reassuring clients about Sanity cloud hosting reliability and self-hosting options

8 replies
Last updated: Jun 9, 2021
Good afternoon all,
I wanted to get some feedback off the back of today's events with Fastly CDN going down (
https://www.bbc.co.uk/news/technology-57399628 ) and causing disruptions across the globe.
I have a client I am trying to convince to head down the route of using Sanity... however they are already not too keen on cloud hosting. I like cloud hosting personally and can boast the benefits all day.

I guess what I'm getting at is how:

a) I will reassure them that cloud hosting is much safer and has less down time
b) is there scope to host your own data with Sanity if they are bullish on the issue

Feedback welcome

Regards

User
AI Update

I understand your concerns - the Fastly outage was definitely a wake-up call for many organizations relying on cloud infrastructure. Let me address both of your questions directly.

a) Reassuring clients about cloud hosting reliability

The Fastly incident context: While the 2021 Fastly outage affected many services globally, it's worth noting that Sanity's Content Lake (where your actual content is stored) remained operational during that incident. The outage primarily affected CDN delivery, not the underlying data storage. Sanity uses Fastly for their CDN layer, but the Content Lake itself has separate infrastructure.

Sanity's reliability measures:

  • Status transparency: Sanity maintains status.sanity.io where you can monitor uptime and incidents in real-time
  • Enterprise SLA: The Enterprise Plan includes uptime Service Level Agreements with dedicated support
  • Security compliance: Sanity maintains SOC 2 and ISO 27001 certifications, demonstrating enterprise-grade security and reliability standards
  • Multiple availability zones: Cloud providers like Sanity use geographic redundancy that most on-premise setups can't match without significant investment

The counter-argument to on-premise: While outages like Fastly's are concerning, on-premise hosting has its own risks - hardware failures, network issues, lack of automatic failover, and the need for 24/7 monitoring. Cloud providers typically have better uptime percentages (99.9%+) than most organizations can achieve with self-hosted infrastructure.

b) Self-hosting options with Sanity

Short answer: No, you cannot self-host the Sanity Content Lake on-premise or on your own infrastructure.

Sanity is a fully managed cloud service, and the Content Lake (where all your content is stored) must be hosted by Sanity. This is core to their architecture and isn't available as a self-hosted option, even for enterprise customers.

What you CAN control:

  • Sanity Studio deployment: You can host Sanity Studio anywhere you want - your own servers, Vercel, Netlify, etc. The Studio is just a React application
  • Data residency: Enterprise customers can potentially discuss dataset region options (EU vs US) for compliance requirements
  • Dedicated infrastructure: For enterprise customers, Sanity has mentioned they can offer hosting on dedicated cloud infrastructure in some discussions, though this still means Sanity-managed cloud, not truly self-hosted

The architectural reality: Sanity's value proposition is built around their managed Content Lake, real-time collaboration, global CDN, and serverless compute (Sanity Functions). Self-hosting would eliminate most of these benefits and isn't part of their offering.

My recommendation

If your client is absolutely committed to on-premise hosting for their content, Sanity may not be the right fit. However, I'd encourage reframing the conversation around:

  1. Data residency and compliance - Enterprise plans can address regulatory concerns
  2. Risk mitigation - Emphasize Sanity's SLA, security certifications, and status transparency
  3. Total cost of ownership - Compare the cost of building equivalent reliability on-premise
  4. Hybrid approach - Studio can be self-hosted, giving some control while benefiting from the managed Content Lake

For most organizations, the reliability, scalability, and feature velocity you get with Sanity's managed service far outweighs the risks of occasional CDN incidents. But if on-premise is a hard requirement, you'll need to look at traditional CMS options that support self-hosting.

Show original thread
8 replies
a) I think if you look at the general uptime for cloud hosted services, and the cost of driving your own hosting, the numbers will often talk for themselves (of course, this depends on what you're really doing).

b) There are no plans of making a self-hosted version of Sanity (because, it's designed to be a globally scaleable real-time content platform). It's like Stripe is probably not planning on making a self-hosted Stripe. But that being said, You are always one export away from getting all of your content as JSON along your assets. And by having structured content, it should by definition be portable and possible to migrate into another system without too much friction (granted, there are probably structures you can do with Sanity that can't be easily expressed in other systems).
You might also ask them what their plans are for doing all the hard work a professional CDN does to minimize the length and severity of an outage. There’s so much work that goes unseen, and if your business isn’t hosting servers, why would you choose to make it a thing that your business also now critically depends on? We don’t see a flaw in cars requiring a safety related recall, and then decide to build our own cars right? We know very well we aren’t going to make that problem any better 😄
AWS loves to say they do the “undifferentiated heavy lifting” and it’s what all big hosting providers and CDN’s do. That work does nothing for your core business and they can do it far better.
user M
I think you've made some great points here. I would completely agree with you. I think sometimes these decisions that people make are purely selfish and all about personal politics. "Why should a handful of big companies hold all the data?"... etc etc. it's tiresome and as you point out they couldn't possibly do it better themselves. I know their answer is to go down the route of a self hosted site developed on a bloated PHP platform which they don't have skills within the team to deal with. I'm just going to give it the hard sell in the near future
Thanks for the input guys
charge extra for “custom infrastructure integration” 😂
The deets are coming out, but this is why you pay someone else to do this. Under and hour to fix a global outage. Can you do that in your self-hosted server when someone “trivially” updates a system dependency that stops mysql from starting properly (etc, etc, etc)? https://twitter.com/User/status/1402627116559077386
👍 I will certainly use this as ammo! Thanks bud

Sanity – Build the way you think, not the way your CMS thinks

Sanity is the developer-first content operating system that gives you complete control. Schema-as-code, GROQ queries, and real-time APIs mean no more workarounds or waiting for deployments. Free to start, scale as you grow.

Was this answer helpful?