You would be really daft not to monitor internet access at all. I have seen members of staff object to certain entities being blocked, but while things can occasionally get fraught it is my experience that more often than not common sense prevails.As Gordon Gekko didn't say* in the 1987 film Wall Street: “Greed is good”. I would add that “B*ll*ckings are better”.Many moons ago I was working with a company which had pretty solid policies and processes but not much discipline – the IT team was great and they did some amazing work, but as is often the case they were a bit informal and gung-ho.It didn't help that one or two of them considered themselves indispensable. Processes continued to be ignored until one week one of the developers figured (wrongly) that I wouldn't notice if he bypassed a security mechanism on the database server, and one of the indispensables behaved outrageously to another member of staff.The former was on the receiving end of a comprehensive b*ll*cking, followed by an explanation of why the rules existed, which to his credit he accepted with grace and from that point was as good as gold.The latter was told that the next attempt would land him in HR under the banner of gross misconduct; when he went crying to management that I had been nasty to him they backed me to the hilt.In most cases then, having the mechanisms to do things right is enough. But when those mechanisms have full management backing and you are seen to act (fairly) on infractions, it soon fixes the minority who think, “that doesn't apply to me”.

No means no Staff engagement is a good thing. Staff who feel that the company cares and gives them stuff or pays attention to their comments and suggestions are generally more inclined to work diligently.Remember, though, that the staff engagement team is generally part of the HR team. You know, the team that makes people redundant, or dismisses them for gross misconduct, or puts them through disciplinary or competency proceedings.The operational efficiency and security of company systems are the first priority of the IT department because they are directly related to profitability.So you are not there to give people what they want; the best they can expect is that their opinions will be sought and given due consideration in the construction of standards and systems.So listen to what people want. Understand what they need. Help them be as efficient as possible within the constraints you have to impose. But remember that the word “no” is often much under-used. Recap The first season of HBO's sharp, satirical Silicon Valley had just eight episodes. By the measure of last night's show, they should have stuck to that for season two rather than try to punch in an extra two episodes.White Hat/Black Hat felt like the episode pulled together from rejected ideas. And there weren't that many, so we were treated to several of them twice. The sharpness was also gone (have the tech writers taken a sabbatical?), leaving some shameful plot devices that making no technical sense.

And as if to confirm that this episode was as unstable as Pied Piper's own compression software, the producers resorted to pointless swearing in the hope no one would notice there weren't enough actual gags. When you reach for the fuck-gun, you know you're not firing on all cylinders.The acting is still good, though: Thomas Middleditch continues to mine a rich seam of comic potential as Richard: failing miserably to down a shot of tequila; doing so again when he decides to give a pep talk to his team; standing up to their ludicrous financial backer Marc Cuban Russ Hannemann...Erlich: Jian-Yang, what are you doing? This is Palo Alto. These people are lunatics about smoking here. We don't enjoy all the freedoms that you have in China. In fact, Erlich (T.J. Miller) gets and delivers all the best lines, and is rapidly becoming the best character in the whole show: savvy yet blinkered, idiotic and brilliant, ideological and utterly shallow at the same time.But before we get to the shameful tech tragedies, here's what happened in terms of plot:

Marc Benioff Gavin Belson realized that his tech giant's competing compression algorithm sucks. Badly. So he tries to hire back the genius who gave us the bionic shit-flinging monkey in order to set him up as the fall guy for Nucleus' inevitable failure. That falls apart though when said fall guy sees the mobile beta and promptly quits. Belson's head is now on the chopping block. Pied Piper is in a bake-off with its only other competitor, EndFrame, to compress 100TB of porn in order to win a critical $15m contract. But Richard feels bad when EndFrame's security guy is fired because they assume his system was hacked when in fact Gilfoyle had found the CEO's password on a Post-It note. This sets up the episode's main plot line: Richard panicking that their system is going to be hacked in retaliation. It doesn't happen. Twice. But in the process, Pied Piper manages to delete hours of its client's "premium footage" and is shown the door by the adult content company. Erlich takes his other startup – an app to help parents find uncrowded playgrounds – to the VCs to get funding. This plot tangent gets the best lines and best laughs as the VCs quickly realize that it is also an ideal app for pedophiles. Erlich responds by using Palo Alto's obsessive antismoking mindset to pivot the app into "Smokation" – helping you to avoid smokers. "No one ever died from second-hand heroin," he parrots back the VC's words. But let's take a look at Smokation. The basis of the software is, we are told, a "geotagging technology that will locate uncrowded playgrounds." The idea is that parents will be able to find playgrounds with few kids. Something that is also useful for pedophiles, it is pointed out, would could provide "some marketing pain-points."

You can see how this would work, in theory. Parents download the app in order to find non-crowded playgrounds. That app provides your location. The app can then figure out where parents are and put them on a map. Simple and wonderfully useless: perfect fodder for Silicon Valley. (There are actually apps out there just for finding playgrounds, in case you were wondering.)But how does that then translate to Smokation? In this case, the location of the people who would benefit most from the app (non-smokers) is not useful; they want location information on people who would derive little benefit from installing the app (smokers). It wouldn't work. This messy thinking is unsurprisingly lazy for a show that has shone by being acutely observed and largely smart in its logic, if tweaked for comic effect.But that wasn't the most offensive piece of tech gibberish. The key denouement is when Cuban Hannemann accidentally deletes terabytes of porn on an external third-party server by placing his tequila bottle on the delete key of a laptop. The networks were connected by FTP and Pied Paper was downloading the files to compress them and, presumably, re-upload to show how efficient and effective its algorithms are.Richard makes the key argument when hauled into the CEO office to explain the deleted data that the speed with which the data was deleted demonstrates the incredible speed of their compression technology. Their competitor would have taken much longer to delete so much data.

We can see one scenario in which this approach may actually happen: for each worker thread, a file is downloaded and has to be compressed before the next file is taken. As such, the speed with which the compression works would lead to the rate of deletion: faster compression, faster deletion. But why would anyone implement such an approach if you were connecting by FTP? You would simply download everything and run compression on it and re-upload.The deletion rate would be determined more by the speed of your connection than the compression rate. And that's not even considering the fact that a company would have to be insane to give a third-party deletion rights on their server when all they are doing is downloading files.The show's tech advisers should have nipped this plot line in the bud or at least come up with a more plausible way for Pied Piper to screw up their potential client's systems. And let's be honest, there are many. So, a bad episode with a few good gags. The season's low point. But with just two episodes left, it's time to ramp things up and there's every reason to believe the show can pull it off.The preview for the penultimate episode shows Belsonioff going full-legal in an effort to kill Pied Piper. Hopefully they will draw some inspiration from the many unintentionally hilarious litigation efforts that have been fired off in the real Silicon Valley.

And that leaves the season's finale where the awkward, amateurish, underfunded and clunky Pied Piper has the opportunity to rise like a phoenix from the ashes and show that the nerds, the dreamers, the good guys can win with some self-belief, mad coding skills and raw, poorly paid determination.As technology evolves, bottlenecks in the infrastructure move around. The switch speed leapfrogs the server speed, then the servers are upgraded with faster LAN cards and the spinning disks in the SAN become the weak link, so you upgrade and find that the SAN fabric is holding you back.How does everything interact? And as the various bits of the hardware keep overtaking each other, can the software keep up or are we simply wasting our time and money?Since we can't afford to have infinite speed in every part of the infrastructure, where do we spend and where do we hold back?What is important is that all the technologies are extremely well established. Perhaps with the exception of Fibre Channel – which we have all found daunting at first but changed our minds once we have played with it a bit – there is nothing I mentioned that most infrastructure engineers find particularly difficult.

The 802.3an (10GbE over twisted pair copper) standard has been around since 2006, and 8Gbps Fibre Channel has been with us a few months longer. So as well as being well understood, they are cheap. And basic 1Gbps Ethernet is regarded as a staple – the laptop I am typing this on has a Gigabit port in it, based on a standard that came along 15 years ago.What is also important in the Ethernet world is the existence of channel bonding – notably LACP and EtherChannel. So if your Gigabit Ethernet network is creaking a little, simply run up a bonded pair and you have doubled your bandwidth with a piece of network string and a couple of commands on the LAN switches and server NICs.The reason why companies have not yet adopted 10GbE for their LANs is that although 1Gbps is not enough, a 4Gbps EtherChannel more than suffices. I have seen a lot of data centres where 10GbE implementations have crept first into the storage network, not the data LAN side.

We have talked about connectivity, now let's talk about the servers you are hanging on the end of the string. Servers have three main elements: processors, memory and storage.I am going to get shot by the CPU police here, but the only fact that I think matters about processors (at least those for servers) is that with each new generation they get faster.Yes, the vendors also strive to decrease the power consumption per CPU cycle, and of course the speed increase is due only to ridiculously difficult innovations in miniaturisation, pipelining, on-board cache and the like. But the bottom line is that processors get faster without really costing more.The story is similar with RAM. As time goes by, you can put more and more RAM into a single slot in a server at a lower and lower cost per gigabyte, but generally speaking the speed of the memory is not your first consideration when you are buying the kit. Quantity is what you look for in RAM.Although there is an overhead of a few per cent of your processing power and RAM if you layer virtualisation on top of the hardware, it is well worth the technological cost because of the ability to allocate resource dynamically and automatically to the virtual servers that sit on top.It means that in extreme circumstances you can give a single virtual machine almost all the resource of a potentially socking big physical host if it needs it for a short while. No need for bottlenecks there, then.

comments powered by Disqus