All 19 entries tagged Work
October 22, 2004
Writing about an entry you don't have permission to view
Matt Chapman has prompted an interesting discussion on his blog; prompted by a remark on Question Time which inferred the only two UK universities capable of competing on the world stage are Oxford and Cambridge, quite a few comments are popping up claiming that this is due to cashing in on historic brand.
I'd like to challenge that view and propose something which suggests that the remark may not only be true but have some substance behind it. If I wanted to assert that Cambridge and Oxford are the only two UK universities able to compete on a truly global scale then how could I do so?
Well, depending on how you rank them, the top 10 Universities in the world are apparently
- Cambridge (UK)
- UC Berkely
- Oxford (UK)
Yale, that's 11th, with Cornell, UC San Diego, and Tokyo next in line. We (UK) next get a showing at 23rd and 25th positions. Then 47th, 60th and 69th.
What is really, really interesting is looking at how these rankings are constructed. The methodology is online here. For those that can't be bothered, these are some things that they don't pay any attention to at all.
- How many work area PCs there are.
- Numbers of type of class of degree awarded.
- Is the Student Union any good
(Hell, they don't even count the employment rates after graduation)
Want to know how to build a University with a worldwide (not just regional, but global) brand in the top ten? Well, this is what they measured and perhaps herein lie the answers.
- Alumni of that institution who go on to win Nobel prizes
- Staff who win Nobel prizes and Fields medals
- Staff who are highly cited researchers
- How many articles published in Nature and Science (they shift this to other areas for places like LSE, which only do Humanities and Social Sciences)
- How many articles published in the expanded Science and Social Science citation index
- Academic performance related to size of Institution
Our very own Andrew Oswald wrote an article in the Independent (which I can't find online) which is how I ran into this site. In it, he suggests that perhaps these are not entirely unrealistic measures for a University to be benchmarked against – our currency is not the investment we make in sports facilities, but our ideas, and these are measures of just that.
Fact is: the world is a big place. If we are going to try and compete on a global scale then we have to understand where we are (and not where we like to deceive ourselves into thinking we are). If we don't, then we won't stand a chance of succeeding.
Just my $0.02 … (pun intended)
October 18, 2004
Having just spent the past hour and a bit assembling our incident management process on the wall for the workshops we're giving in ITS over the next few weeks, we now have the side of our meeting room looking something (or even exactly) like this,
I feel slightly like I've just been through a protracted episode of Blue Peter. Give me a badge!
August 31, 2004
August 26, 2004
So, we might barely just have a reliable internet connection back.
Tuesday night, at around 10pm, bad and evil things started to happen to our link to the Internet – or rather to the link between our firewall and the site router. It's out of term time and the side effect of this was that no-one reported anything to the emergency number that evening. This meant I woke up on Wednesday morning to discover a distinct lack of Warwick presence on the net.
After having a momentary panic about what was at fault (hard as it is to believe, it could have been much worse) it was pretty clear something was up with our firewall. Sure enough, there it was, clocking up approaching 90% CPU utilisation. We used to have this years ago on our old 100Mbit throughout firewalls, but this one has 8Gbit/s of throughput (our link to the 'net is only 1Gbit/s).
Evidently something wasn't right. It didn't take very long to see a vast amount of incoming traffic on highly dubious ports – much of which was to a small number of systems that had been turned off the previous day due to virus infection. Best guess is that there was some form of delayed 'dial home' attack underway from machines which these had themselves compromised.
It got better, it got worse. We got medieval on incoming traffic, which helped for a short while, then it went again. It wasn't until 8pm last night (by which time the three of us still working on the problem had decided there wasn't anything left to do but replace the firewall in the morning – cue sound of straws being grasped) that things then returned to normal, albeit with limited incoming traffic.
Anyway; we decided to swap the firewall out with a replacement this morning (better safe than sorry) and have continued today to review,revise and remove firewall policies that were permitting what we think is unsuitable traffic in. Net result of this is that we still have services not accessible (particularly some in departments, which is a real nightmare for us, since many of those are as important as our own) but most things are up and perversely, in terms of incoming traffic, things have probably never been as well protected as they are now.
A busy day awaits tomorrow: the two days of my week that this has consumed were supposed to be occupied with, amongst other things that have slipped, writing up a job description for a new post in time for a Hay evaluation panel in early september – for which the deadline for submissions is tomorrow. Joy.
July 25, 2004
Writing about an entry you don't have permission to view
So far, so good.
Sounds like we have an idea now why the generator failed; also, very positively, we have managed to keep power going to the phones and network through the UPS and generator whilst the mains was being worked on.
It's around 7:00pm and we're now starting to bring services back up. No air conditioning yet in the machine room (bit warm in here!) or power to our building, but that's all in hand.
Plenty to do still, mind you.
July 21, 2004
The one thing I find restrictive about the blog service is the layout of some parts of the admin interface. The button which takes you 'back' to the previous page (either the main admin interface or from there to your blog) is either at the top of the page or at the bottom. When you click on it at the top you then have to drag the mouse down to the bottom to click it.
It would be much nicer just to be able to click-click and not have to move the mouse – I really notice this on my Macintosh with a trackpad.
July 14, 2004
Boy. That was yet another HighlyEnjoyable™ afternoon in IT Services. Sitting in someones office on the ground floor of IT Services when at 3:40pm the lights go off in the room and a spluttering sound is heard from the standby generator, which is co-incidentally just outside the window. A spluttering sound which, rather worryingly, rapidly ceases and which is replaced not with the expected, re-assuring hum of a generator, but with deafening silence.
T-7 minutes later, with the generator still not working, our UPS finally runs out of power (only being designed to provide enough to allow the generator to kick in – that is, after all, why we bought it) and all power is lost to the ITS machine room on central campus, wiping out probably three quarters of our central services, including access to the Internet, a huge chunk of the University's telephone network and numerous other systems too many to list here.
From an entirely (and admittedly rather selfish perspective), I take some small consolation that the network itself recovered once power was returned, with the only exception being our site router (responsible for Internet access) where corruption caused by the uncontrolled shutdown needed to be manually recovered. Small consolation indeed though, since there have been no shortage of other problems resulting from this fault in need of extensive work.
Those of us still remaining on site (thanks guys!) finally wrapped up everything we could deal with by around 8:30pm; the problems which remain outstanding (two of the filestore servers, one student and one staff groupwise postoffice) are so serious that the repairs which were started earlier will take most of the night and early morning to complete and the earliest we'll be able to progress with them is tomorrow morning.
Worse, we don't know for sure that there aren't other problems waiting in store for us, currently hidden by the ones still being fixed, and assuming that the fixes underway actually work! So, we still have approximately 1/3rd of our staff GroupWise users without email access – including most of us in IT Services – and a fair chunk of staff who won't have home directories when they log in first thing.
Oh, and no-one can tell us for sure why the power failed (something 'tripped' in a substation, possibly due to an overload, although it was apparently not only ITS, but also the Ramphal Building, parts of Physics and Chemistry affected). Great.
Here's hoping that the morning is slightly less HighlyEnjoyable™.
July 09, 2004
The Bethesda Fountain in Central Park is one of the most famous sights in the park. Situated within Bethesda Terrace, the fountain contains the only piece of sculpture commissioned as part of the original design of the Park. "The Angel of the Waters", as the sculpture is known, commemorates the purification of the citys water supply through the construction of the Croton Aquaduct in 1842, which bought fresh water to all New Yorkers. Before this, the rapid population growth that occurred in the 19th century resulted in the existing city water supplies becoming polluted and unsanitary, resulting in increases in diseases and in epidemics of cholera and yellow fever.
The fountain & sculpture, in particular the angelic figure, are based upon a story dating from the time of ancient Israel. The Pool of Bethesda in Jerusalem was reported to be bestowed with healing powers by the periodic (annual) visits of an angel who stirred the waters. The first person to enter the pools after each visit would be cured of whatever ailments they suffered. The pool itself has long dried up, but legend has it that on the coming of the new millenium, the waters of the pool will return and all those who enter them wil be cured of any illness or disability.
At the base of the statue are four smaller figures; these represent temperance, health, peace and purity.
Anyone who watched the TV production of "Angels in America" recently (show over two nights on channel four) should recognise the fountain – the final sequence was filmed at Bethesda Terrace with the fountain as the backdrop for most of the piece.
June 30, 2004
Well then, that hurt (again) today. Frankly, given what a rough time we've had with the network so far (both users/customers and staff within ITS), I'm suprised that I don't have nightmares about people coming into my office at 8:40am asking me "do you think there might be a problem with the network?".
Today was, in one respect at least, thankfully far removed from the nightmare we had on the Thursday during the last week of the summer term. On that day we had two problems, the second one of which could have been induced by the former, which massively complicated troubleshooting. It was also rather gutting to have got through most of the term with a particular software feature disabled until it could be fixed, which itself would have prevented the second problem from escalating in the way it did. For the sake of 48 hours that day would simply never have happened.
By way of contrast, there wasn't a hard fault – induced or otherwise – with the network today. Instead, what turns out to be the case is that at some point on the network (which we managed to isolate to a relatively small number of possible locations on the residential side of things) someone, somehow, managed to create a 'loop' on the flat campus LAN. We've had this problem happen before – in fact it's been a running sore for many years, but this time the problem was slightly more subtle. We've got used to troubleshooting hard-down events where a switch becomes a 'smart piece of wire' and forwards traffic incorrectly down links which should be shutdown by default, doing so at very high speed and swamping the network. We've done a huge amount of work to mitigate against this (of which the patch for the software was the last piece of the puzzle).
What was different this time is that the network itself wasn't swamped. Links were running at well under 10% capacity, no sign of problems on the switches. It was actually symptoms reported to us by our systems team that pointed us in the right direction. This time, it wasn't running out of bandwidth that was the problem, but rather that on that part of the network, the loop was causing massive amounts of confusion. Devices didn't know where to forward traffic to, the routers were seeing traffic from themselves, and generally things were not happy. On the other, routed, portions of the network (including University House) things were far more stable provided you weren't trying to communicate with anything in the affected part of the network.
The real killer unfortunately is that in the affected part of the network are 90% of our centrally provided servers. So, those users in routed parts of the network (like Uni House) who depend on access to those servers were, despite being protected internally, still very badly hit by the disruption.
The lesson we can take from today is that this wasn't (thankfully) a failure of any of the equipment, as such, and that as we start to migrate servers into our new machine room in University House, and onto new equipment in IT Services, the situation will improve for those people who are already routed. Add to that the migration work we'll be able to now do over the summer to move the large departments (such as WMG) into their own routed networks, and we'll finally be starting to use the capabilities of the new network to prevent problems like this from happening, instead of spending time firefighting problems which are due in no small part to our having to maintain the legacy network structure until we reached a time where we could continue with the migration process.
The one small grain of comfort I take from events like this is seeing how staff in ITS react when serious problems occur. This isn't often visible to people outside the department, but the way in which staff through the department will pull together and focus on solving the problem is, during incidents such as this, what keeps me sane. You all know who you are; thank you.