London-based market data consultancy CJC has launched a cloud-based managed service for Refinitiv’s Data Access Control System (DACS) platform, at the same time confirming a new technology partnership with Google Cloud aimed at supporting CJC’s overall cloud strategy. By making DACS in the cloud a viable proposition, CJC hopes to eradicate entitlement management as an obstacle to wider cloud-based delivery of market data services.
CJC already provides cloud-based managed services for multiple market data solutions, including Refinitiv’s TREP (which DACS is a part of), BCC Group’s One Platform and Push Technology’s Diffusion Server. CJC also manages 150 traditional DACS implementations globally.
For the past 30 or so years, DACS has been used by TREP customers to manage user entitlements, and to enable proper control and auditable records of which users and applications are entitled to publish data through the TREP environment. The new DACS in the Cloud solution from CJC is aimed at both existing and new DACS users, and is being offered at varying levels of service, from simple cloud deployment with remote management, to a fully managed, technical and commercial solution.
Easing customers’ pain
According to Peter Williams, chief technology officer at CJC, DACS has created significant challenges for organizations over the years, particularly around the demands of upgrades of operating systems and hardware. For example, because of the way DACS is structured as a database, most upgrades require the schema to be updated as well, which can expose new tables for new functions or open up different areas, he says. Another challenge for firms is the global consolidation of regional DACS platforms. Many global firms run a DACS instance in each region they operate in, which can result in multiple IDs, for example. If information from the three versions is brought back together, the data needs to be normalized.
In fact, says Williams, the idea of offering DACS in the Cloud emerged out of conversations with existing clients, many of whom required quite a lot of work to update their DACS instances. Clients have also indicated that it was a good first step on the journey towards working with more cloud-based solutions, he says.
Williams says that CJC has taken advantage of cloud-native technologies, including Kubernetes, Virtual Machines (VMs) and Containers, to speed up deployment and make ongoing management easier. He says CJC can put a customer which is on version 6.4 of DACS and deploy a 7.3-level version of DACS in the Cloud in a half-hour. He also makes impressive claims for maintenance. “Once it’s deployed, we can actually allocate more resourcing under the covers,” says Williams. “So if an organization, let’s say like a small hedge fund, deploys a DACS and it’s for 16 users and that business grows, we could take their existing DACS and under the covers grow it to be a far bigger system, and that would be effectively seamless to the user.”
As well, he says, “we add our own management and maintenance elements to it, so we can maintain, we can monitor, we can audit and make sure that the controls are in place. We can make sure that it’s locked down, who has access, what they have the capability to do once they have access. It’s a real leap forward, really, in the terms of how things are done and for the end user, simple. For us it’s simple because all the engineering and hard work was done up front, and it’s been wrapped into the tooling.”
Looking to next steps
Having completed the DACS project, Williams says CJC is also looking at other opportunities to offer cloud versions of solutions. “The buzzwords for us have been around simplifying or how simple can we make this complex thing, and acceleration and enablement,” he says. “How can we deliver these things fast for customers, how can we help them achieve their goals quickly and in good order?” For example, CJC has been working with Google Cloud for a number of years, and became a Google Cloud technology partner in mid-February as well.
Williams says that it will be some time before there are Cloud solutions for algorithmic trading or other processes that require ultra-low latency. “But that’s not the largest part of the market. You know, you’re talking the 80/20 rule, that’s probably the 95/5 rule,” he says. “What we’re saying is that the people that dismiss market data in the cloud because it can’t do those workloads are kind of missing the point.”
Equally he says that those in the industry who think that at the moment the cloud can only handle reference data, with five updates an hour, are underestimating the cloud’s potential. “We took it about how far along the curve can we go,” says Williams. “At the moment with Google, as part of our partnership, we’re actually testing real time streaming market data, and some of our initial test results are showing sustained update rates at half a million updates per second with mean latencies of about 250 microseconds. That’s enterprise class market data delivered.”