Private cloud completes Cloudera's data platform vision, but it has turned a corner into even fiercer competition

Where the big dogs roam

Cloudera has introduced support for the private cloud - built on the Red Hat OpenShift platform and supported by Kubernetes container orchestration - to elbow its way into a whole new world of competition.

The erstwhile Hadoop-shifter reckons the move will let customers run data services from a single integrated platform via public, private and hybrid clouds. It will be able to funnel data into reporting, analytics and machine learning that draws from enterprise applications, data warehouse, desktop systems and external data sources.

The data platform would also offer the same security, privacy and governance across cloud instances, Cloudera claimed.

Arun Murthy, chief product officer, told us that extending the environment across all cloud instances would "accelerate time value and ensures critical workloads meet their SLAs".

If the system works as advertised, it means users can manage their data in a private cloud but spin up workloads in the public cloud where necessary and appropriate, from within the same platform.

Adam Ronthal, Gartner research VP for data and analytics, said Cloudera's private cloud support might appeal to organisations yet to fully embrace the public cloud model due to security, governance, or data sovereignty concerns.

Other providers are also playing to this market, including Oracle Cloud at Customer, AWS Outposts and Azure Stack, he said.

"Cloudera starts from a position of on-premises strength while the cloud services providers are obviously strong in their own environments. Which one of these is more appealing is going to depend on where an organisation's centre of data gravity is today – cloud or on-prem – and where they expect that to be in the next two to three years," he said.

The decision would also be based on which vendors the user organisation is prepared to commit to in the long term. To start with, Cloudera's hybrid cloud data platform is most likely to appeal to eisting punters and those that need a hybrid cloud strategy, addded Ronthal.

"It provides perhaps the easiest lift-and-shift from an on-premises approach into the cloud and it provides cloud portability. It is generally best for end-users that have an existing strategic commitment to an independent software vendor's (ISV) offering. On the other hand, it may require a little bit more effort to extend that into the surrounding cloud service provider's ecosystem, and it represents an explicit decision to select a point of vendor lock-in. You're locking into the ISV, rather than the cloud service provider, but no matter what you do, you're going to have some degree of lock-in," Ronthal said.

Bigger pond

But Cloudera could struggle as merely a smaller component of a broader analytics landscape, where it's going to be competing with cloud-native approaches. Cloud vendors are already expanding their catalogue: there's Azure Synapse analytics, as well as cloud-native Hadoop products like Amazon EMR, GCP Cloud Dataproc and Microsoft HDInsight, each of which has their own tight integration points with their surrounding cloud ecosystem.

Cloudera's fortunes have been improving since it published disastrous Q1 results in 2019, after Hadoop – the big data file system it was built around – went out of favour and it struggled with the merger of former rival Hortonworks. The first quarter for fiscal 2021 ended 30 April saw revenue of $210m, an increase of 12 per cent as compared with the first quarter of fiscal 2020.

The improved performance reflected better management of the combined company, and growing acceptance of its Cloudera Data Platform. Whether it continues on this path will depend on how it fares against a whole jungle of ISVs competing in the same space, some fairly new to the market, such as Dataiku or Talend, and some with a deep history, like Informatica.

All these vendors are trying to get products in a sweet spot of meeting the users' multi-cloud data management needs beyond the services built into the cloud vendors' environments.

"What these third-party approaches are trying to do is push on the line of 'good enough'," said Ronthal. "When the users buy into that, it is because they feel can get additional value from an independent offering."

But the competition in the data platform market is fierce, with a long list of vendors vying for users who want more than "good enough". Meanwhile, the cloud infrastructure providers have deep pockets and continue to invest in their data platform products. Cloudera has turned a corner, but it may yet find obstacles in the route to fatter sales. ®