Service providers think their operations and billing software (OSS/BSS, for short) is inadequate for orchestrating the coming wave of networking, and they’re eager to find something new — and more automated — as a replacement.
That OSS/BSS is antiquated isn’t a surprise; these are complex, often proprietary software systems that built up inside telcos over the decades. It’s well known that there’s some attic-cleaning to be done. But the need is becoming more urgent as software-defined networking (SDN) and network functions virtualization (NFV) come closer to reality.
The Metro Ethernet Forum (MEF) has begun pitching an OSS/BSS successor called Lifecycle Service Orchestration (LSO). It would start off as a catch-all layer, sitting above the OSS/BSS silos to act as a single conduit of communications to/from the services layer.
A study on LSO from The Rayno Report quantifies carriers’ urgency. Issued during Mobile World Congress earlier this month, the report includes a survey of 60 service-provider professionals about the hopelessness of traditional OSS/BSS (which, from this point, we’re going to abbreviate as just “OSS.”)
Fifty-four percent of the respondents called their OSS systems “outdated,” and some are placing bets on an alternative already. In five years, 20 percent of OSS spending could be shifted toward LSO, analyst Scott Raynovich concludes in the report. (Some carriers think the figure will be a lot higher, he adds.)
By 2019, the total LSO market, counting LSO itself and OSS migration to LSO, could total $2.75 billion, Raynovich says. And the rampup could start soon.
“As SDN and NFV get deployed this year and next year, that’s going to set the stage for LSO to get deployed,” Raynovich said during an MWC reception at the CENX booth (pictured above, with Raynovich at right). “People want to use these standards to prove it out.”
OSS handles a wide variety of tasks that could be described as the operations back-end of a telecom provider’s business, handling dozens of tasks such as provisioning, alarm monitoring, and billing.
In many cases, those functions were built in a per-service basis. That is: When a new service was created — such as Carrier Ethernet connectivity, or a new VPN or security offering — it often came with its own OSS to be bolted onto the rest of the OSS machine. Even better, OSSs tended to be proprietary or customized. As telcos acquired one another, many of these OSSs were left operational, even if they duplicated each other’s work, because merging them was too enormous a task.
This is part of the reason why carriers can take weeks to provision services. There’s a lot of software to be manually navigated.
Operators hope that LSO can lead to automated and fast provisioning — even self-service provisioning. In addition, LSO, or anything else that modernizes OSS, could handle real-time tracking of the network and automated responses to alarms.
Some of this has been discussed on the SDN side, where vendors say the network could eventually monitor itself and make adjustments for traffic changes.
LSO is the MEF’s idea, but it’s theoretically an opportunity for OSS vendors. Carriers are looking to those vendors to flesh out the LSO concept, Raynovich said. The implied alternative, assuming carriers really are that frustrated with OSS, is that they would go find someone else to do it.
Regardless, it’s clear that the major service providers have a vision for where OSS should go.
“Several large service providers have indicated to us that their goal is to open up and change the way the OSS systems interact with the applications in the data center. It only makes sense that, if more of these services are deployed in an open SDN and NFV model, they communicate with OSS systems,” Raynovich writes in his report.
The DevOps Angle
Service providers told Raynovich they’re working on cultural changes as well, trying to think and behave more like cloud service providers.
That means embracing a DevOps style — continually revising code, rather than perfecting it before release, for example — and opening up networks to let the developer community create new services.
“We believe the developer community will start coming up with creative ideas to use our infrastructure. We think that’s a critical piece,” James Feger, VP of network strategy with CenturyLink, told Raynovich for his LSO report.