If you’re on the Microsoft Azure cloud and you need a blazing fast search engine for your application, it’s likely you’ve already considered running Apache Solr on Azure. The open source search engine is one of the most widely used and full-featured search platforms available. But getting Apache Solr setup and then maintaining and managing on Azure all by yourself can be be tricky and time consuming. You must make a decision between building, managing, and supporting your own custom Solr infrastructure in Azure OR using a trusted partner who can run and manage Solr as a Service Azure.
Do you want to accelerate the development of your feature-rich search applications on Azure with a Solr-as-a-Service provider? If so, we’ve put together this list of the 8 biggest things you need from a hosted Apache Solr Service on Azure.
Requirement #1: Effortless, Fully Automated Management
One of the biggest benefits of a fully managed cloud search service is the automation services you’ll receive. If you need more time building your search applications and less time managing, maintaining, and monitoring them, the entire provisioning and scaling of Apache Solr Search on Azure can be managed for you. This means quickly and easily spinning up single nodes – for your development, QA, or UAT environments – OR multi-node high availability clusters running behind load balancers for your production environments. Service that can offer on-demand or scheduled backups, ability for you to restore from those backups, ability to add nodes on the fly, log archiving and management, 24/7 monitoring and alerting, and analytics— all automatically completed every day. With a Solr-as-a-service provider, you can automatically deploy a Solr cluster in Azure in minutes.
Requirement #2: Native Solr, Without the Restrictions
One of the drawbacks to some Apache Solr-based search providers is that they have built an API layer on top of Solr. This essentially forces you learn their APIs, train your team on those APIs and above all change your code base (indexers or search APIs) to now deal with with these vendor APIs. To minimize this, make sure you find an Azure Solr Service solution that lets you access Solr natively. If you ever need to switch providers, the switching costs will be low – just point your application to the new Solr URL’s.
Requirement #3: High Availability is Essential
You’ve built a robust search application that serves your customers well. Any significant downtime is disastrous. Leave the hard work of building highly available Solr search clusters in Azure to the Solr and Azure experts. Don’t be on the hook for high availability. An Azure Solr as a Service solution allows you to focus on the critical requirements for your business.
Requirement #4 – Disaster Recovery
Running in cloud environment requires a different kind of thinking. Even the biggest cloud providers have shown that entire data centers can be impacted, so planning for disaster prevention and having a DR plan in place that allows you to meet or reduce your RTO and RPO objectives is critical. Looking for a provider that offers Disaster Recovery in the region of your choice, whether that’s Active-Passive or Active-Active is a very important consideration.
Requirement #5 – Security and Compliance
Security and compliance requirements are seemingly growing by the day. If you have an internal search application or provide search over sensitive data, you need security whether it’s on Azure, hosted on premise, or any other cloud provider. If security is important, you must have the latest security patches and updates applied on a regular basis. What’s your (or your current search provider’s) process for security updates and patches? How do you apply Apache Solr security vulnerability patches to your infrastructure?
For applications that demand high security, you can opt for your clusters to be not available or accessible via public URLs. A higher security and compliance option is to choose a private virtual network. These types of deployments are hosted in their own Azure Virtual Networks (VNets and are connected via VNet peering). This isolates your search infrastructure from other applications and other traffic. It is a strong security measure that guards against attacks, DDoS and other threats that can compromise your application security and compliance commitments.
Does your Azure Solr-as-a-Service provider offer you audit logs? At anytime, can you go and check who’s accessing your system, what action they are performing, and when and from which IP they came from?
You’ll also want your Azure hosted Solr platform to be GDPR compliant as new requirements around data protection and security are implemented. Are they Privacy Shield Certified? Ask for their Data Flow Diagram.
Requirement #6 – Monitoring & Alerting
You want cost optimized Apache Solr on your terms, but without the responsibility of managing every detail and being “always on”. Once your cluster is built and running the way you want it, you’ll need to implement a monitoring system. Your Azure Solr-as-a-service provider should provide functionality to alert you on the health of your servers and cluster as a whole. Can you configure specific alerts around the health of your Apache Solr clusters? What if your Cache Evictions are taking a toll on your cluster’s performance? Can you get alerted on those?
Requirement #7 – End-User Search Analytics
It’s critical that you have comprehensive end user search analytics as part of your solution. Leverage native application-level API’s to track all user interactions going on in your search applications. These API’s are always listening and recording so you can gain insights around: performance, relevance, conversion, and more. These signals help your search team to fine tune and optimize your search results for a better search experience.
Requirement #8– SLA-backed Up times and Support with Credits
A bit of a no brainer, you’ve made the decision to move your search infrastructure to Azure. Make sure you get the peace of mind that strong SLA’s and warranties for breakage are in place for your Azure Solr Service solution in terms of both uptime guarantees and support response times.