Logo - Full (Color)

HubSpot Reliability

How we keep customer data safe and build a platform that's reliable.

How does HubSpot make its system reliable and resilient?

We build our software platform so that HubSpot is available and accessible in a variety of disaster scenarios. Each service in our micro-services architecture is spread across multiple servers running in different data centers. Our services use API's to communicate with each other to reduce the interdependence between each other. Every service also has a corresponding test environment where changes are deployed before they are migrated to production. 

HubSpot also utilizes a worldwide Content Delivery Network (CDN) to distribute content to a location closest to users enabling quick and consistent access wherever you are.

 

Are HubSpot’s site and services always available?

Our goal is that you can always access your HubSpot account. There are times when the HubSpot service will be unavailable due to planned maintenance or due to a component failure. In such cases, HubSpot staff are paged as soon as the failure is detected and work to make sure the service is back up in the shortest possible time. You can check our Status site for the latest status of our services.

 

How does HubSpot make sure outages due to component failures do not reoccur?

When an outage or significant failure occurs, HubSpot's primary goal is to get the service up and available to customers. After the issue has been resolved, the team that owns the affected service holds a post mortem, i.e. a formal review of the incident. During this review, teams use “The 5 Why’s” process to analyze the root cause of the event and to develop a list of immediate action items to make sure this event, and other events like this one, do not re-occur. Action items are expected to be completed within 30 days of the event. Each post mortem event is documented in detail and future learnings are incorporated into long term plans.

Is my data safe with HubSpot? 

The HubSpot platform uses a variety of datastores to store data and ensure data safety. Each datastore is architected using best practices for data safety and recovery. HubSpot products are hosted with Amazon Web Services and Google Cloud. Data stored in the HubSpot platform is replicated to three data centers. If a server in one data center fails, the processing is switched to a replica server in another data center with minimal service interruptions.  HubSpot also maintains hourly and daily backups for each data store, these backups are maintained for 30 days. Backups are maintained on highly durable media. Backups of a critical subset of our customer’s data are also maintained in a different geographic region to protect against a regional disaster. 

 

Is my data secure with HubSpot? 

All communications between a web client and HubSpot servers are protected using TLS (1.0, 1.1, 1.2) protocol encryption using 2048 bit keys. We also provide customers with the ability to enable Two-Phase Authentication (2FA) to prevent unauthorized use of their portals. Communications between HubSpot services are protected by using Virtual Private Networks and encrypted network protocols. Data is encrypted at rest to help protect against unauthorized access.

What is the geographical location of the site where my data is stored? 

HubSpot data is primarily stored in Northern Virginia, USA. A critical subset of data is also backed up to Ohio, USA. Additionally, we have some data that is replicated to Frankfurt, Germany. Additional details on our cloud infrastructure can be found here.

 

How does HubSpot ensure all my data is backed up and can be restored in case of a disaster?

The disaster recovery strategy at HubSpot uses a combination of snapshots of data and daily full backups to ensure that there are multiple copies of data available to be restored. Snapshots are designed to provide a quick recovery mechanism where the recovery can happen in minutes. Full backups are used when snapshots are not available to recover the data.

The microservices architecture of the HubSpot platform enables operations staff to prioritize getting high impact services like a sales rep’s ability to get work done or getting web pages available before attending to other tasks like the ability to publish new pages.

 

Can I recover my data if I accidentally delete it? 

Based on the context of the deletion, it is possible to undelete and recover data within a day of an incident or the data being accidentally deleted. Note that if you use the GDPR contact deletion outlined here, this is a permanent delete and we cannot recover this data. Full-service subscription customers can enter a support ticket at help.hubspot.com for our staff to determine the best way to recover your data.

What does HubSpot do to monitor its systems?

HubSpot operations and engineering teams use industry-leading tools and instrumentation of services to monitor and analyze the behavior of our SaaS platform. Metrics from services and our cloud infrastructure are fed into an alerting framework. Alerts generated from our alerting framework will cause our automation to take the appropriate corrective action or it will notify operation staff of an abnormality that needs review and attention. 

 

We’re seeing reports of website downtime through an external tool, how can we resolve this? 

A website can be unavailable due to a variety of reasons. Check our Status site to determine if there are any issues with the HubSpot service or if a regional internet outage that may be causing an issue. If the Status site does not show a current issue, please open a support ticket at help.hubspot.com

 

How does HubSpot let its customers know and keep them updated? 

If we find issues that might affect your ability to use the HubSpot service, we will post it immediately on the Status site. Updates to current issues will also be posted on the same site. We have also built notifications within your HubSpot portal that will notify you of functionality that may be experiencing issues at that time, look out for a banner that will explain the affected service.