BLOG

Like many that have been in IT for as long as I have, I’ve become very skeptical of the “hero numbers” that vendors promote on their website or in social media to get attention.  Most professionals know that these hero numbers would NEVER be achievable using real world conditions. Our conditioned response is to immediately call BS because we’ve all been conditioned to seeing hero numbers posted in big bold letters and immediately look for the asterisks and qualifiers written in fine print (‘your mileage may vary’) at the bottom of the page.

When presenting the Vexata solution to customers, partners or analysts, I can understand why the initial response is that the VX-100 specifications are “too good to be true”.  The skepticism is understandable – I get that, but I’d confidently put this system against ANYTHING in the market, regardless of cost – because the results speak for themselves.

From an integrity standpoint, Vexata stands behind the specifications that we publish 150% and to prove that, we put our money where our mouth is (so to speak).

To that end, Vexata felt it was important to get third-party validation on our system performance in Oracle environments, so I reached out to Brian Garrett at ESG Lab to discuss a potential lab validation. IT pros value the integrity and credibility that ESG brings, mainly because the ESG team really knows their stuff and they won’t publish it if they can’t validate it. I met with Brian and Scott Sinclair at the Flash Memory Summit in Santa Clara and walked them through the Vexata story and they were visibly excited (drooling?) to further engage. Brian connected me with Tony Palmer during our initial kickoff call the excitement on both sides was quite apparent. We scoped the project, and the details of the VX-100F NVMe Flash and VX-100M Optane Storage Systems and were all very eager to get started.

Why is this kind of validation so important?  More than just validation of speeds, feeds and boring specifications for throughout and IOPS which typically causes eyeballs to roll, this validation addresses a critical business problem regarding consistent, predictable application latency under load.  Consider the amount of enterprise capital that goes into database licensing (often priced per server core) every year. Now imagine the thousands of cores deployed across any enterprise data center to support operations and analytics, it’s a huge expense in any CIOs budget. Now imagine explaining to your CIO that you are only able to utilize 30-40% percent of those licenses because of I/O bottlenecks, not exactly a great career move.  So, any solution that allows you to achieve better utilization of those licenses would make you the hero that delivered immediate ROI and lowered the overall TCO.   This is “The Bigger Truth”.

Don’t get me wrong, the performance numbers that ESG reported are astounding and trust me when I tell you that these are NOT hero numbers.  These tests were performed with the utmost scrutiny using real-world workloads and benchmarks (SLOB, TPC-C, TPC-H and Hammer DB) recognized by Oracle architects as the “real deal”.  ESG Labs characterized the performance of both the VX-100F (Flash) and VX-100M (Optane) systems for IOPS and latency under heavy workloads across multiple I/O types and sizes.

The heaviest system stress was when ESG placed the system in a HTAP (Hybrid Transactional and Analytical Processing) workload, simultaneously using SLOB (Silly Little Oracle Benchmark) to drive an OLTP workload and using Hammer DB to drive OLAP analytics against the same 5TB Oracle database. This test would crush any mortal all-flash array, but Vexata has built something much more than that.  The test results shown on the Vexata dashboard below speak for themselves.  If that wasn’t enough, the team also verified these results using Oracle Cloud Control to validate that host performance matched the storage system analytics.

Vexata Dashboard: HTAP Workload – OLTP, Data Ingest and Query Processing

Vexata will put these performance numbers against any system in the market – regardless of cost.  How can a startup be so bold, you ask?  Because these are the exact same conclusions that our customers have reached – and they typically vote with their wallets.

Customers also know that application infrastructure is measured by much more than just performance thresholds and business metrics, but it’s also about the ability to deliver a system that scales on demand, provides resilience and fault tolerance against any number of potential failures scenarios.   To validate this, ESG embarked on a number of “Enterprise Ready” situations that focused on the management, fault isolation, analytics and ability to handle hardware failures, configuration changes and software upgrades without disruption. This is truly where the Vexata VX-OS architecture really shines.  ESG captured a number of different failure modes and upgrade scenarios while the system was under heavy load and captured all the resulting data.

I’ve personally been involved in a number of these lab validation reports with ESG and while all have been good experiences, this one is by far the most rewarding.  It’s clear that Vexata provides a high level of transparency to our partners and customers and this comes through on this report.   Please take a moment to download and read through the report and provide any comments – our goal is to educate and inform. That’s the Bigger Truth.

You can download the report here.