Anodot provides ‘Autonomous Analytics’ as a full SaaS product.
We also support a managed service approach to cater to customers with unique security challenges.
Our managed service (on-premises) has several limitations and it is inferior to our SaaS product.
ON-PREMISES PRODUCT LIMITATIONS
- Up to 200k Metrics. If needed, more metrics are available in an additional instance,
however, this instance is separate from the first in terms of metric, users, dashboard
alerts and the like.
SLA and Access
- Remote access is needed in order to collect metrics and in order to manage the service.
- Remote SSH Access needed for troubleshooting and to manage the service.
- A new version is built and shipped once a year. Version update is mandatory.
- Bug fixes, if required are built on top of the current version.
- Feature set is limited. Data Collectors are not supported.
ANODOT INSTANCE REQUIREMENTS
See instance requirements in a dedicated article - Single Machine System Requirements
Question Any requirements for high availability specific to the Anodot application?
Answer Not relevant. All is stored on the same machine. - All in one.
Question Services included/excluded in the on-prem implementation by Anodot.
Answer Everything but
- Data collectors
- S3 access is required for images in Webhook & Slack.
Question My understanding from some previous communication is that the host OS can be something
other than Ubuntuas long as we have kernel 4.4.x. Can you confirm this
and provided a list of supported OS if applicable?
Answer We currently only support ubuntu 16.04 LTS
Question Does the Anodot on-prem application container reference a base OS
image or does it rely on the host OS?
Answer Yes. rely on host OS
Question Application network requirements?
Answer Ports to be exposed/mapped to host and allowed through firewalls:
- For managed service support
- 443 for monitoring
- For internal use
- 8000 for UI
- 8080 for APIs
- Both should support HTTP and HTTPS
Question Application scaling guidelines so we can understand how it’ll scale out and what the
underlying system requirements are.
Answer Does not scale up. Limited to 200K metrics. All in one. Customers may create several instances
which are not connected.