I like to commute to Infinio HQ in Kendall Square, Cambridge, in the company of friendly and familiar voices. I listen in to the wit of The Cloudcast, news from Speaking In Tech, or the technical banter of shows like vSoup and VuPaaS.
Today’s listen let me be a fly on the wall while Bob Plankers and GS Khalsa spoke about infrastructure and some of the complications they’ve seen in the wild. One statement struck me enough to tweet it:
Bob is speaking to the complexities of ensuring VMware SRM is not just configured correctly, but will achieve the business objective when you need it to do so. GS has a great story to tell around minute 23:00. The conclusion, without ruining the surprise, is that the testing did not test like production was running.
What struck me this morning is how relevant this conversation is to the testing my customers have done with Infinio. I have at least one customer who installed Infinio, had a good first impression, but just didn’t see much with regards to results. The customer did not see much because there was barely a workload on the system — no I/O most of the time, with sporadic blips going up to 100 IOPS.
The customer decided we had earned his trust by making such a thoughtful product, and that we could install directly into production — where latency was already an issue.
Lo and behold, we saw results.
What is worth learning from this story? Should we test all new software right in production?
I am not convinced all products deserve a chance at jumping the queue to production. What I do know is that Infinio requires no new hardware, given you have some resources available to us. We install without interruption to data. And we uninstall with a single click in the UI, again, with no interruption to data.
Only we can offer you all these benefits to a trial run.
Why not run with it? I say let's see if a product this simple can meet your needs before you worry about qualifying all kinds of complex alternatives.
Matt is a Sales Engineer at Infinio