Guido Vrola - Fotolia
"I never had doubt this company was going to make it," Violin Systems CEO Ebrahim Abbasi said of the all-flash...
array vendor that emerged from bankruptcy six months ago.
If true, that means Abbasi may be the only person who never questioned Violin's viability. Even Violin customers thought the company was dead, Abbasi said.
And now, with Violin Systems back in action, it still faces a tough battle convincing people in the storage world that it can carve a place for itself in a competitive market.
Following a slow decline into bankruptcy, the George Soros-backed Quantum Strategic Partners investment group rescued the all-flash array vendor in April 2017. Abbasi, who served as Violin's COO under CEO Kevin DeNuccio, convinced the Soros group to give Violin Systems another chance and took over as CEO. Abbasi previously served as president or COO of Force10 Networks, Roamware and Redback Networks.
Now that Violin Systems is debt-free, Abbasi predicts profitability by the end of 2018 and plans to buy a software company to complement the vendor's Concerto OS capabilities. Like other flash vendors, Violin is also preparing to ride the approaching non-volatile memory express (NVMe) wave.
Violin is selling the all-flash arrays it revealed but never got a chance to launch before bankruptcy. Still, it faces an uphill battle with nearly every storage array vendor betting their futures on flash arrays. Those competitors include Dell EMC, NetApp, IBM, Pure Storage and Hewlett Packard Enterprise (HPE).
Abbasi filled us in on his plans to save Violin Systems, one of the pioneers of all-flash storage arrays.
What does Violin Systems look like today? How many employees do you have?
Ebrahim Abbasi: We still have less than 100 and 80% of them are engineers. We went down to less than 40 and built back up. We are investing significantly in research and development. We rebuilt our engineering and customer service. We'll go step by step and focus on how well we take care of our customers versus how many customers we have left.
All of our competitors have more customers and revenue than us. Success for us is not going to be determined just by technology, but also by how we respond to our customers, with care and attention and responsiveness.
It's a team effort, and I have assembled a team. I can assure you everyone around me is smarter than me.
What are your financials like now? Violin was never profitable, either as a private or public company. Can you make enough money to survive?
Abbasi: You may be surprised to hear that, for first time in the history of the company, we became cash-flow positive in the last quarter. We're headed toward being profitable. I told Soros' team, I'm going to make this company profitable; I don't want to go from hand to mouth. I don't want to get into trouble from the beginning.
When will you be profitable?
Abbasi: Let me put it this way: It will not be beyond next calendar year. But to give a specific date would not be prudent.
What does your product portfolio look like now?
Abbasi: We have a 140 terabyte platform in two versions. One, the FSP 7650, is a nondedupe system that gives 2.2 million IOPS and less than 1 millisecond latency. The other, the FSP 7450, provides 6-1 to 8-1 dedupe ratio.
We have revitalized the platform focusing on the way the old Violin was at the start. We used to be able to go to a customer and say, 'I'll fix your performance issues.' We provided that, and the company thrived.
Your main flash arrays now are the ones you launched last year just before bankruptcy, right?
Abbasi: This is part of the sad story of this company. Those two models that came out last September were the first two products that came out of the engineering and R&D that I reorganized. We announced in the market, and days later announced we were going into bankruptcy. So, we never introduced them [until after bankruptcy]. We have since added significant features, and we are now shipping with replication, encryption, deduplication, compression; complete data services, all developed by Violin.
We fixed the dedupe issues on our operating system. We also took the operating system and loaded it on an x86 server, enabling it to recognize any pool of storage. We unhooked the tie up between the operating system and the Violin system.
We have more than 120 customers. Now, we're focused on our next steps.
Ebrahim AbbasiCEO, Violin Systems
What are those next steps?
Abbasi: We want to provide software-defined storage to high-performance users so they can manage storage more cost-efficiently. We're also adding tiering and scale-out capabilities. To do that, we are looking to purchase a company with advanced features in that area. We want to finish this on our own, so we are taking elements of our current operating system, elements of buying it from someone to speed up development, and we'll finish the rest for the next version of our operating system sometime next year.
I don't expect you to tell me who you're going to buy, but when do you expect to complete an acquisition?
Abbasi: First, I have to do my homework. I don't talk about things until they're done. I'm just giving you a vision statement. We are looking at a couple of companies. It will be a small purchase.
Have you added 120 customers since bankruptcy, or does that include prebankruptcy?
Abbasi: These are not all new customers; some are returning customers. Customers started looking elsewhere because they thought we were dead. Now we're visiting them, they're coming back in droves, from Fortune 50 to Fortune 100 customers. During our bankruptcy, customers were not even returning our calls. Now, they're calling us.
Are they buying new systems or still using their old arrays?
Abbasi: We have customers who are buying our new platform, customers who are upgrading their old platforms by adding more capacity, and some customers come back just for upgrades to the latest software for functionality.
The flash array market is extremely crowded and competitive. Before bankruptcy, your competitors made sure potential customers knew Violin Systems was struggling. What are you competitors saying about you now?
Abbasi: I don't care what competitors say about us. I care about what customers want and how I can give them service.
One customer said one of our competitors told him, '[Violin] won't make it. Soros bought them and they don't have experience in storage.' I said, 'We're not actually in the business of selling storage. We're in the business of selling experience.' The market is big enough for all of us. I don't waste my time bashing Pure, [Dell] EMC, NetApp, [HPE] or IBM. My focus is what problem can I solve for my customer?
When customers talk about our competitors, I change the narrative. My competitors are not bad. I've learned a lot from our competitors. We're going to provide the best solution for our customers that we can.
Everyone selling flash is talking NVMe. What is Violin's plan for adding NVMe solid-state drives (SSDs)?
Abbasi: We will deploy 3D NAND and NVMe. We are testing now. NVMe is ready today, but it's a question of when customers are going to use it and in what way. NVMe brings two things to the market: one is latency and one is cost. SSDs are becoming so much more powerful, but customers have to make changes in their operations. Cost of change is prohibitive. I think there is an optimistic view that the market is ready for NVMe. In my humble view, it is not. I think it will be 2019 before NVMe is fully deployed, and, by then, we'll be ready.
How did the Soros group get involved with Violin Systems? Did you talk to them before the bankruptcy sale?
Abbasi: Soros had a pre-IPO history with Violin. Then they were part of the bond that my predecessor raised. They were the largest bondholder. It was a race between the largest shareholder and the largest bondholder, and the largest bondholder won.
When the company went through bankruptcy, because I was running the company, they came to me and said, 'What are you going to do with this company?' I said, 'I'm going to maintain what's working and what gave us thought leadership, and expand on the software side, because focusing on hardware only is not a winning proposition.' I gave them a strategy and they bought into it. They asked me to stay, and we moved on.
Pavilion Data, Vexata enter NVMe flash array market
What's the true cost of flash?
Who needs an all-flash data center