28 Sep 2020

GMA Networks

GMA Network (Global Media Arts) is a major national commercial broadcast television and radio network in the Philippines.

Its first broadcast on television was on October 29, 1961 and is headquartered in Quezon City.

GMA is a major national commercial broadcast television and radio network in the Philippines and a major Asian television leader. As such, they are extremely active content creators, producing many of their own programmes as well as huge volumes of content for delivery to multiple platforms.

GMA was one of the first to make the business decision to further extend its leadership role by investing heavily in in-house 4K production. However, it knew that its existing SAN system was already struggling to cope with the demands being placed on it to generate the content it prided itself on producing in-house, let alone the demands of 4K productions.

Out with the Old

The performance of GMA’s SAN had been acceptable for many years, but had clearly become outdated. It was servicing a small workgroup of approximately 30 operators, which is less than half of what GMA wanted to meet its business objectives.

Although GMA had a lot of experience with SAN and had not quite given up on it, it was pretty sure SAN wasn’t part of its future. That begged the question, “What is?”

So, GMA went looking for new storage.

Looking for the New

GMA drew up a comprehensive tender to accommodate the large workgroup it wanted, i.e., more than 65 client connections on 10GbE, plus a range of 1GbE connections for less demanding tasks. In total, somewhere between 100 and 120 users – many of whom would be working in 4K – would have simultaneous access to the system. This was the foundation of the system by GMA, with the overall bandwidth started at 10GB/sec and wound up at 15GB/sec.

Image

Real World Testing

The GMA Requirements

Drop Outs.

As one of the tender responders, GB Labs soon noticed that other manufacturers were falling at the early hurdles, largely because of the tender’s detailed requirement for a “true” Disaster Recovery strategy. i.e., properly separating master and slave, which almost none of them could demonstrably provide.

Except, that is, GB Labs and one other, who were the only two companies who stayed in the race, both prepared to quote for the project and provide proof that it could do both storage and disaster recovery at the levels required.

But that race didn’t last long.

Living in the Real World

GMA, rightly, insisted on real-world testing by both companies, not only on the proposed healthy storage system, but on the system as it was failed.

GMA was astute enough to consider what many others fail to do, that is, understand that when a cluster-based system has a node go wrong, the entire file system has to start the process of reoptimising, which can take days – or longer – during which time the system performance becomes seriously degraded, limiting usage.

The competitor system proof-of-concept NAS was quickly exposed as not being able to achieve anywhere near the performance GMA required, even in a healthy state, and when it came to failover testing, its performance was even worse, failing both tests badly.

On the other hand, when GB Labs’ proof-of-concept was submitted and tested, and failover testing was also completed, the system performance remained rock solid in both. It didn’t even waver.

Image

Failover

In just two seconds

GMA’s decision was immediate and crystal clear, saying, “There is only one system that can do what we want for our expanded, 4K production operations. And it’s also demonstrated that when it fails over, it won’t stop, or even slow down, which is absolutely crucial to our workflow. We have got to have GB Labs.”

The reason disaster recovery is so crucial is explained in the following and is directly tied to the amount of content GMA wanted to generate.

Protection guaranteed with CloakDR

Generating massive volumes of content means there is an equally massive responsibility to protect and preserve it. Therefore, failover is crucial, and the system GMA had been using had not only failed to live up to expectations, but was proven incapable of handling the number of people GMA needed to be working from the system.

A solution had to be found because, when disaster strikes, it’s already too late to talk about disaster recovery.

Failover tends to be overlooked by less informed facilities, even though it’s one of the most crucial components for guaranteeing system availability around the clock to meet production and broadcast deadlines.

Fortunately, GMA is not one of those facilities. During the induced failover aspect of the competitive system testing, the negative impact that losing a node had on the incumbent system had zero eect on GB Labs. GMA was savvy enough to know, despite claims by competitors of the infallibility of a cluster approach, that it needed a true DR system from someone who truly understands that failover is essential.

That’s precisely why GB Labs developed CloakDR, which blankets a system with an impenetrable, protective disaster recovery cloak that oers true, demonstrable resilience and redundancy.

And “resilient” doesn’t mean CloakDR will “do its best”. GB Labs designed and built CloakDR to provide no-compromise redundancy by enabling dual GB Labs systems to act as both master and slave. Unlike other systems, GB Labs storage is one hundred percent live. Data written to it is not confirmed until both slave and master have it, which keeps both in-sync

The benefit of that is obvious. If the master suddenly vanishes for any reason, CloakDR ensures that the slave, at a separate location, takes over with exactly the same IP address and paths as the master in less than two seconds. No other system can match that, and certainly not the one GMA previously had in place.

Yes, other systems can recover from the unexpected, but it’s usually a lengthy, painful and often very expensive process that can take days, which still doesn’t take into consideration the loss of revenue due to downtime, or the loss of precious content, which can take hours, days, even weeks to recover.

But in CloakDR, GB Labs provides what is expected in terms of redundancy, and that’s resilience, speed, and most importantly, confidence. GMA had exactly those expectations. It was easy to switch GMA from its existing system to a fully operational, safe, and secure GB Labs CloakDR protected system.

Failover GMA

Summary

The speed, resilience, and failover capability of any GB Labs storage system is unparalleled, and those benefits are particularly important to other potential GB Labs storage and CloakDR users in the region. With the proven success of GB Labs systems at the PyeongChang 2018 Olympic Winter Games and Paralympics, including large workgroup 4K editing, and the 2020 Tokyo Olympics and Paralympics rescheduled to 2021, it is a crucial time to consider how to reliably protect and retain content – at any resolution – particularly in a region known for its dedication to the highest standards of quality, speed, and resilience.

In content creation, production, and delivery, the more you have, the more you stand to lose, but there’s everything to gain if your facility’s performance is accelerated and protected by GB Labs.

It’s the clear winner.