PAU L OCKENDEN
Paul offers a friend some IT guidance, and looks at the downsides of straying from the big four mobile networks
Paul offers a friend some IT guidance, and investigates the downsides of straying from the big four mobile networks.
About a year ago, a mate of mine asked me to do an informal audit of the IT systems of a local company where he’d just been hired as head of technology. What I found can best be described as a colossal mess. The server room had around 30 servers of various makes and vintages, some in racks, but many just piled on top of each other. All of the machines were switched on, consuming lots of electricity and creating an earsplitting amount of noise.
The servers were all plugged into unmanaged switches via a bird’s nest of tangled network cables, none of them labelled. Many of the machines were also plugged into domestic four-way extension leads! There certainly weren’t any power management devices to be seen.
The whole thing was a shambles – and, more importantly, a disaster waiting to happen. Obviously, I won’t name and shame my friend or the company here. What I will say is that it’s an outfit dealing in tech electronics and really should have known better! Said company isn’t alone, though. Over time, many firms find themselves slipping into a similar state.
My friend and I eventually managed to log into each of the servers. There was no central management, obviously, although the Windows servers were at least sitting within a domain, and the various Linux and other boxes all shared the same root password as the Windows administrator account. Not a good thing – but it did make our task much easier!
We found that of the 30 or so servers, six were powered up and doing precisely nothing. Three didn’t even have network cables attached. And of those that were doing stuff, four were just churning away on jobs no longer needed. So we whittled things down to approximately 20 useful servers.
The first step in sorting out this mess was to suggest a degree of consolidation. There were several file servers and around ten web and intranet servers running various applications. With the aid of an Excel spreadsheet, and a bit of juggling, I was able to reduce the box count to just 12.
But then I added some more. Why? Well there was no redundancy at all in the firm’s systems. It had an SQL server running lots of important tasks, but it was just a single instance. So, I suggested that as a bare minimum a second SQL Server be installed with mirrored databases, set up within a high availability group.
Likewise, the intranets and various mission-critical web applications were running in a standalone mode, so I suggested secondary servers with automatic failover in case something were to go wrong. Same again for email. And we needed to add a backup appliance, too. We were quickly back up to around 20 servers again.
Now, had this been a normal firm I’d probably have been suggesting that much of this processing be moved to the cloud. Tools such as AWS and Azure are pretty reliable and very economical. But because of
“I struggle with VDP. I’ve found it to be unreliable, needing a lot of nursing to keep it going”
the particular industry this company works in, and the nature of some of its contracts, use of cloud computing and storage is strictly forbidden.
Twenty physical servers is daft, though. The key to sorting out this mess was virtualisation. A single, beefy server would probably run virtualised copies of all of these physical servers quite happily, but then we’d be back to having no redundancy. If the big server goes bang then you end up with a whole company twiddling its thumbs.
So for reasons I’ll explain in a minute, I suggested three physical servers plus a SAN for storage (a good SAN will have redundancy built in). Also, high availability pairs of managed network switches and PDUs, to avoid a single point of failure. It’s an ideal setup for a SME – and the whole thing can easily fit into a half-height rack, leaving plenty of space for other comms kit, routers, firewalls and the like. In fact, it’s pretty much a reference setup for SME virtualisation.
Why three servers? Well, I like to have two running all the missioncritical services with mirrored data and files. Items such as intranets, web servers, email and database servers. Each physical host can run half of the “live” services to even out the CPU and memory load. In the instance that one of this pair of servers dies, the other can continue running everything.
The third server is used for development, test VMs and anything non-essential, and also a temporary repository should you need to move a VM off one of the main servers.
When it comes to running a virtualised environment on these servers, there are plenty of options. I’ve tried most, but always come back to VMware. If you’re a real cheapskate then you can get away with running its ESXi hypervisor on each server for free (albeit with limitations), but I’m a fan of VMware vSphere Essentials kits. These are specially designed for a three-server setup such as the one described above. You can have a maximum of two CPUs in each server, but with today’s multicored server CPUs, this isn’t an issue.
There are two versions of the Essentials kit. The standard one costs a shade under £500 and gives you a full-fat copy of the ESXi hypervisor for each machine, plus a copy of vCenter for centralised management of your VMs. If you can stretch to it, though, I’d recommend the Essentials Plus Kit. It has the same three servers and six CPUs limit as the basic kit, but Plus bundles a shedload of useful software. It costs just under four grand, but I think it’s well worth it.
As well as the hypervisors and vCenter, the Essentials Plus kit offers a high availability function that restarts a VM on an alternative host should the main one fail. It helps to keep everything running should a physical server fail.
Then there’s vMotion, which I love. It lets you migrate a VM from one physical server to another while it’s actually running! I’ve tried setting up a continuous ping while this happens and you might see a fraction of a second outage, but that’s all. All the software running on the original server just continues to run on the new one. Session state is maintained in any applications, and even file writes just continue from where they got to before the migration kicked off.
In my own company, we use vMotion to move VMs to chunkier hardware if there’s a temporary load increase (perhaps a web server getting swamped because something has gone viral), and then move it back again a few days later. vMotion also means you can easily move all the VMs off a host if you need to do hardware maintenance or update any firmware, without any downtime whatsoever.
The VMware Essentials Plus kit also offers Data Protection or VDP, which is a backup appliance based on EMC’s Avamar technology. It offers agentless backups and has built-in deduplication – but I struggle with VDP. I’ve found it to be unreliable, needing a lot of nursing to keep it running. Backups should be “set and forget”.
I’ve recently discovered an alternative product called VM Backup from Paragon. It’s designed to work with the VMware Essentials kits (either the standard one or the Plus Kit), even having the same three-server/six-CPU limit. It’s another £344, but well worth it. So far it has been working well for me, but I’ll report in more detail in a future column.
We haven’t finished with the Essentials Plus Kit yet, though. There’s also vShield Endpoint, which provides agentless antivirus and antimalware protection for all of the VMs running on the three hosts. Finally there’s vSphere Replication, which provides asynchronous replication between VMs at hypervisor level. It will work across sites, but is also good at keeping backup copies of VMs ready in a multiserver setup. It works well with the high availability software I mentioned a few paragraphs back.
The other bit of the jigsaw is VMware Converter, which is free, and will take a running server and create a P2V (physical to virtual) clone of it, suitable for running on a hypervisor. Again, other tools are available, but I’ve always got on well with VMware Converter.
The only thing you need to be aware of with the VMware Essentials Kit is that there’s no way to go beyond the three-host/six-CPU limit without paying lots of money. Essentials is designed for SMEs, and VMware has deliberately imposed limits that protect its more expensive “big company” revenue streams.
But with modern servers, you’d be amazed at just how much you can consolidate onto three hosts. Especially when the VMs will be copies of ancient servers that probably have little requirement for disk space, memory and CPU grunt.
So what did the company do? Well, as I mentioned, my audit was an informal one. I was expecting it to call in “proper” consultants to tell them pretty much the same thing. Instead, it went with my Excel spreadsheet. The company spent a few quid on new hardware and software, but saved a fortune in electricity and air-conditioning costs. It has also reclaimed the old server room as a new office and the halfheight rack now sits in a broom cupboard, making hardly any noise!
Virtually identical?
I guess most people are aware that here in the UK there are really just four mobile networks: EE, O2, Three and Vodafone. Even these big names share base stations and parts of their infrastructure – so the real number of networks will average out at somewhat less than four.
All the other mobile networks you’ve heard of are MVNOs – or Mobile Virtual Network Operators – which piggyback on top of one of the four core networks. So, for example, Virgin Mobile uses EE, as does BT (which is hardly surprising as BT now owns EE). Tesco Mobile runs across the O2 network, as does Sky Mobile and Giffgaff. Again, this makes sense once you realise that O2 and Giffgaff share the same parent company, Telefonica. Vodafone doesn’t have many MVNOs on its UK network, but two that you might have heard of are Lebara and Voxi. And then there’s Three, which is the network behind the ever-popular ID Mobile and up-and-coming Smarty.
You’ll nearly always find that the MVNOs are cheaper than the big four networks, and often by some margin. Sometimes, they offer extra goodies such as data rollover and caps on out-of-plan charges. So this makes it a no-brainer – everyone should sign up with an MVNO over one of the big four networks, right?
Maybe. But maybe not. You see, things aren’t always quite as simple as they first appear. Just because an MVNO runs on a particular network, doesn’t mean it gets everything that the host network offers to its customers. Indeed, it doesn’t even mean that the MVNO’s customers get the same network coverage.
As an example of this, take Plusnet Mobile. It uses the EE network, and like its host network, Plusnet is owned by BT. Plusnet Mobile even boasts within its advertising that it’s “Proud to use the EE network”. There’s just one problem – Plusnet only uses part of EE’s network.
EE runs its 4G services at 800MHz (Band 20), 1,800MHz (Band 3) and 2,600MHz (Band 7). Band 3 is used for the main core of the network, while Band 7 is used to provide additional capacity in cities and other built-up areas; it’s also the band that EE uses for its “double speed” 60Mbits/sec service. Band 20 is used primarily for Voice-over-LTE services – normally a phone will drop back to 2G or 3G when a call comes in, but VoLTE avoids that. It makes sense to run this at 800MHz because that’s better at penetrating inside buildings and other structures, and works well over longer distances such as rural locations.
That all sounds great, and for EE customers it is. But Plusnet Mobile uses only Bands 3 and 7. The MVNO has no access to EE’s Band 20 services, so this means that there will be places in the UK that you’ll get a signal on EE but not with Plusnet. I asked the company about this, and it said “we’re working on the ability for Plusnet customers to access 800MHz so they can get even broader coverage in the future”. That’s fine, but I suspect most people reading the “Proud to use the EE network” claim will assume that the maps are identical today.
Don’t get me wrong. I’m not singling out Plusnet; I’m using it as an example because I’m an actual paying customer! If you look at other MVNOs, there aren’t only frequency bands missing, some of them don’t even offer 4G. And even when 4G is offered you’ll often find that, with an MVNO, the data rate is capped at a lower level than the host network, and the traffic will be given a lower priority.
It’s hard to get actual figures for much of this stuff, especially since there are a couple of websites out there claiming to carry definitive data but that are actually putting out misleading information. But going back to Plusnet again, I found a comment on its support forum from one of its technical staff saying that the data rate is capped at 40Mbits/sec, whereas the EE host network offers 60Mbits/sec as standard, and 90Mbits/sec on “Max” plans.
Also bear in mind that the main four networks tend to offer Wi-Fi calling where the handset supports it, or VoIP apps where it doesn’t, to help in areas where coverage is poor. Most of them also provide official (and legal!) signal booster boxes. Such things aren’t unknown in the MVNO world, but they’re pretty rare.
Now you might be prepared to live with these limitations in exchange for a huge drop in your monthly fees. I certainly am; I have contracts with both Plusnet and ID Mobile. But I think it’s time these MVNOs stopped giving the impression they offer the same coverage as the underlying network, when in many cases they patently don’t.
“An MVNO won’t necessarily get everything the host network offers to its customers”