Wednesday, 22 April 2009

Suffering Support!

A while ago, I upset some of the senior managers – after a particularly stressful day, I suggested that we retire all of the existing office staff, and then go down the local junior school and select 30 children at random. I offered my opinion that these young people would be more IT literate than our current employees!

Of course, I was joking (mostly); but sometimes, it is difficult to understand why people find the simplest things so hard to do. For example, if you can’t print, then it would seem a fairly simple thing to do to check that the printer is turned on, and has paper and ink in it. Yet not a week goes by without someone complaining that their printer is “broken”, and then upon investigation, we find that it has run out of consumables.

A few years ago, I tried to carry out some research into the effectiveness of our support team: I analysed the number of support requests and how quickly they were resolved. But then I looked at what the actual items were; and most were what we would refer to as very basic IT problems. Some related to simple hardware issues such as cables pulled out, others were minor software related items; where is my “lost” file etc? Only a few really needed “an IT person” to fix, just someone with a reasonable level (as we see it) of common sense.

Now the first reaction was to check if this was down to laziness on the part of the staff member; after all, let’s be honest, it is easier to pick up the phone and call for help than to try to fix something yourself. However, it became obvious from the analysis that there was a large number of support issues that came down to the various individuals lacking some pretty basic IT skills.

I then carried out some more research and it soon became clear that many of the staff had actually had no real computer training – at best, they had been shown a sequence of steps to perform; click button A, click button B then button C, then print. But if it didn’t print, they often just pressed the same buttons again as they didn’t know why they had to press those buttons. When I asked, they had difficulty explaining what they were doing. They just couldn’t explain in terms that I could relate to (in many cases because they didn’t know); and some of the instructions I offered were clearly just as meaningless to them.

This worries me; a great deal of time is wasted and productivity suffers as a result. Users feel frustrated that they are unable to work with the tools, and morale begins to suffer. Then the IT staff turn up, and too often the user feels belittled by someone that shows annoyance at having to deal with a very minor problem.

So the immediate reaction would be to suggest that all staff should be better trained; clearly this offers benefits, yet too often, the reaction of management is that the users don’t need to become “IT trained” in order to carry out simple tasks. This is possibly true – but equally, we cannot continue to put people in front of a PC and give them so little training that they make as much work as they actually achieve.

To offer an analogy, many people drive cars – but we don’t allow people to climb in and drive off without undergoing some form of training and taking a basic skill test to prove they are ready to be let loose on their own. Whilst using a computer is not quite as serious (for the most part), it still makes sense to ensure that the person using it has a certain level of knowledge to make them as efficient as possible.

Tuesday, 14 April 2009

Printing for Fun and Profit

When I started with my current company, I was more than a bit surprised at the number of printers - almost every other PC had a printer attached, and most of those were inkjet printers. There were so many different makes and models that keeping track of them was a problem - and there were almost 40 different cartridge types that had to be bought. We were ordering replacements every single week, and not just a couple at a time, but literally dozens. We had a storage cupboard just for these cartridges.

Now to me, this is simply crazy. Although there are some inkjet cartridges that hold larger quantities, most will have less than a standard pub spirit measure (25ml) - and while you can still get a single measure of scotch for £1.50, the same amount of ink will set you back £15 - £25. Given the choice, I know which one I'd prefer!

My predecessor had simply bought cheap printers - and every time, he bought a new one, it needed a different type of cartridge. He also considered it easier to attach them directly to a PC so there were very few people doing any form of network printing. The problem is of course that this is exactly what the printer manufacturers want - they sell the printers at a cheap price, knowing that they will make their money on the ink (and boy, do they make their money). And of course, a cheap printer will fall apart quicker, but hey that's OK as it doesn't cost that much to replace it (although you'll then have yet another different type of cartridge!).

So we started by working out where the most appropriate place was for a networked printer, then ordered up some decent mono laser devices. Within a year, almost all of the older inkjets were gone. In fact, we now have about 50% more users, but 30% less printers. And we are now down to about 5 toners with about 5 inkjets, so it's easier to keep track. Plus, the laser toners last much longer so it works out much cheaper. In fact, we reckon that we save about £25,000 to £30,000 per year over what we had been spending.

Now of course, there are those that will say "use compatible cartridges" or even "refill them". Tried those and they are an absolute waste of money and more importantly, time. You end up spending so much time fiddling around - I really think we have more important things to do. Also many of the compatibles don't work; and a lot then cause cleaning issues. I just think that it's far better to get decent equipment and be done with it.

That's not to say that users were happy; far from it. When we first started moving them onto networked printers, you would not believe the fuss it caused. People really don't want to walk 20 feet to get their print jobs. But eventually, they started to accept it and now the situation is so different. Although we still get some problems, there are far fewer than we used to have, administering them is easier and most of all, we save money that we can then spend on other things that help us do the job better.

In fact, this is such a simple thing to do, that I am astonished how many companies are still using large numbers of inkjet printers. I spoke to a manager at another company and they have around 500 users - he told me that he has one member of his IT staff almost permanently dealing with printer issues. He couldn't actually tell me how many they have as they have actually lost track of them; they think that they have over 300. (that's not a typo; yes it's three hundred!)

Now no disrespect to him, but for me, that is a red flag - time to make some serious changes. And if people complain, then I would insist on charging them for buying cartridges. If these consumables end up on their budget, I guarantee they will take notice!

So do yourself a big favour - get rid of all inkjet printers, replace them with laser printers. You'll feel better for it.

Thursday, 9 April 2009

Email me

I do actually remember sending my first email. It would have been around 1990 - I had been playing with PCs for about 3 years. I had managed to get some odd bits & pieces of equipment from an auction; one of these was a modem that ran at a blistering 2,400 bits per second. And in those days, there was no plug and play, you had to create files that would contain the necessary commands to make the hardware work.

In my case, it actually took several days to get it working, but finally I managed to get it to dial out on command. The only problem then was what to connect to - there were not the ISPs we have nowadays. I actually managed to connect to a Bulletin Board Service (BBS) that I think was based in North London; I remember watching with fascination as the screen suddenly displayed a list of numerical options - of course no GUIs in those days. I vaguely remember that to post a message was option 5; I sent some messages to a couple of people, but there weren't that many people you could mail.

Of course it has all changed now; email is a major tool of the business and it is difficult to imagine how we could work effectively without using it. Every department relies heavily on being able to contact others, both inside and outside the company. And unfortunately, this leads to some major problems. Previously, the company had very few effective controls on the email and no email policies at all. I made my self quite unpopular because I insisted on setting these controls in place and worst of all, enforcing them with total rigidity.

To begin with, we upgraded to Exchange 2003 from 5.5 which was quite a task in itself - nothing had been prepared and the AD wasn't quite right. Later we had to move from an older server to a newer rack mounted device with more processors, memory and storage, and that process was much easier - in fact one of the easiest migrations I've ever had to do. We also started to get people moved from all different varieties of Outlook onto one version (2003); it took several months, but we got there.

Among the changes we made was to implement a fixed limit on mailbox sizes - just 200 Mb. Now for most people that is not a problem; but we regularly keep an eye on the sizes, and every time, it's the same names that appear close to the limit. I regularly get asked to increase the size limit but refuse; if I increase it, they will just let it run to the new higher limit, then complain that isn't enough.

To provide some assistance, I set-up space for mail to be archived to a .pst file on a designated space on a server, so it gets backed up as well. People have been shown how to archive, but you could be forgiven for thinking that we hadn't as so many users still don't seem to be able to do this.

We also limit the size of attachments on incoming and outgoing mail to 5 Mb. You would not believe the number of emails that get rejected due to the size of attachments - and the size of some of these; we still get complaints that a mail won't go through when it has a 20 or 30 Mb attachment. We've shown pople how to use other methods of transferring files, but they just don't want to know. As an example, I checked the mail logs yesterday and there was another incoming mail rejected from a user at our parent company in Germany - the person had attached a spreadsheet of 14 Mb, and sent copies to 5 other users in Germany, and 6 in the UK. When it failed, the person then re-sent it to all 11 people twice more!

Still, I suppose that I can't complain; the system is now working really well (apart from the users!) - we have even managed to add the domain for our French company to the system and they now connect to us for their email. Il marche tres bien! In fact, it looks as though they are particularly impressed as we fix issues for them far quicker than their previous host.

Just now have to wait for Exchange 2010....

Friday, 3 April 2009


At my previous company, the network cabling was pretty poor – it was put in by a guy they referred to as “Dodgy Dean”. I’m told that he is wanted by 3 separate police forces and is currently living somewhere in the Costas!

Basically, it was a mix of old coaxial cable (with BNC connectors), some Cat 5e cable most of which done correctly, but some of which was reversed (green / orange). It was run through gaps in the walls, through drains, over the roof, through guttering, mixed in with power cable, under motors – pretty much a mess. There was absolutely no structure to it what so ever.

The network used to crash throughout the day – tests showed that quite a number of the cables were dead (around 20%). I also found some strange configurations in the hubs, with cable crossing over and packets travelling unnecessarily longer distances. The speed was pretty poor (as you might expect) and people would often accidently cut through or disturb cables to cause disruption.

I eventually decided enough was enough – I set about replacing all the crap. As I started to pull the old cable out, I found even more hidden away than I knew about – it turns out that he never took anything out, just ran more cable when one piece failed. Some of the old cable was covered in all sorts of green, brown and black slime! Much of it showed signs of the outer sheath cracking or breaking up as it was in such poor condition. In the end, I filled a whole standard skip up with the dead cable and still had enough over to half fill another.

Eventually, it was all tidied up, documented and tested and the network became a lot more stable and transmission speed improved. Then I changed jobs!

When I started with my current company, it was a bit like taking a step back in time. Again, there was a mixture of older coax cable, some fibre and mostly Cat 5e. The cable runs were really poor, just shoved in where they could with no planning of any real kind.

So I trained up one of the staff, showed him how to do basic Cat 5 ends, and punch down in the patch panels. I spent some time analysing what was really needed through the offices and factories, then considered how best to replace all of the garbage.

We started with some fairly basic changes – moving some cables and putting some new ones in a more controlled way. Then we added proper basket tray above the office spaces to carry new cable and it started to take shape. We always run cables in threes – the extra cost is minimal, but it provides a lot more capacity and it is easier than trying to add in later.

When the company decided to build new offices, it was agreed that we would cable it up – we had had quotes in the range of £38,000 to over £50,000. In the end, we put in almost 10,000 metres of cable for the new offices and the cost of materials was a little over £3k. Add in another £3k for staff time involved and it became a major cost saving. It actually allowed them to plan for a really fancy lighting system that cost some £30k.

In the comms rooms, the patch panels look really pretty as they are very carefully patched. In most places, these normally resemble explosions of spaghetti, but ours are very different and nicely laid out. We have had several people in from outside companies that are astonished at the quality of the work we have done. (See the picture above for an example of a comms rack - we have 3 full height units, 1 half size and 2 wall cabinets, and they all look like that.)

So why do we put all this effort in? Well, we take a pride in our job; we want to have something that we can point to as high quality work. It also makes our lives easier; if there is a problem, it takes less time to identify and resolve. The structure we use provides much more capacity for growth with minimal effort – it’s also a lot easier to move people around when necessary.

The problem is of course that within the company, most people don’t recognise the effort that has gone into it. All they are interested in is that the system works when they want it to, and as a result they just don’t appreciate that not everyone takes as much care as we do.

Thursday, 2 April 2009

DPM Update 2

Having sorted out the problem with the autoloader, we can now add the tapes in and it displays them on the screen - it shows the barcodes on the label attached to the tape to help keep them ornganised as well which is quite useful. However, one small issue - 2 of the tapes are marked as "suspect". This is strange; they are both new and have only been used for a basic test.

It turns out that this is a known problem and there is a fairly simple solution. Details can be found here It's to do with the database used by the software - it marks the tapes with data and this can get conflicted if the data and the label don't quite match. It appears that this is what has happened to us as the tapes were used for the testing, before the bar code label was put on.

A couple of caveats for this process; there is a script that you should download and add to a text file to run as a command. Make sure that the script is all on a single line - if it isn't, then the script fails. You might also need to watch what AV solution you use as some don't like you running command files.

It is also important that you follow the described process; remove all suspect tapes, run the script, then add just one of them and run the script again. Then after it detects the tape correctly, erase the data on the tape. Repeat until all tapes are added and data erased. If a tape failes, take it out and add it to the end of the line and use one of the others, until all tapes are OK.
It took a while to complete the process - nearly a full day, but we got there in the end. So all is well....

Well not quite. Now the reporting services function doesn't work. Again, it's known problem, but most of the sites that offer fixes for this, only do so for Server 2008 and we are on 2003. Still, I have hopes that we will get a fix for this fairly soon.

Decided to edit this rather than post a new item having spent a few days working with the guys at Dell. Really helpful chap called Javier who is a SQL specialist - he determined that the reports ran OK through a separate browser window, but not through the DPM console. From this, it appeared to be an issue with the .ASP settings and he did eventually get it fixed. Unfortunately, he didn't give me the exact details as he was working on it using a webex session for several hours. However, from what he told me, there are some issues with access rights - it's probable that the update carried out earlier reset those, which is why it wouldn't run.