Penny Wise...
One of the most touted advantages of "new technologies", the "new" as in "new" kind and the "new" as in "ain't new at all" kind, is the fact that you can save money by using them. At least you think so. The idea is that if you use computers to do something, is gonna take less time and so you can get more money... Somehow... If you use this... Hummm... Thing, to do ... Stuff... You can do it faster and it will make you more money! So the plan is: 1. use new thing to do old thing, 2. ... ? 3. Profit!
Of course everybody is quiet on the fact that to do 1. you first need to purchase the 'new' thing, and sometimes the thing is expensive, a lot. And then the point 2 is a bit vague. And point number 3 is also foggy.
If you use a computer to do something, are you actually saving time? And are you actually going to make more money? Sometimes, actually, all the time, I don't think so. Everyime I start doing something with a computer after a while I realize that I've spent an inane amount of time lost behind small details, details that aren't that important. If I am writing code on paper, I'd probably adopt a different approach. I'm not saying it would be a better procedure and even a faster one, but it could be.
What is anyway clear, is that there are a lot of peoples out there ready to sell you something that is supposed to do miracles for point 1 and is very good at making vague gesture about point 3.
In this sense is obvious that the best thing to do is to spend as little as possible at the beginning and check later if the not-spent money are actually really saved or not. In many cases the answer is no. So spend a little in the beginning is better than getting nothing later. Assuming that you can still do your job.
But this practice (spending as little as possible in the beginning) is not always the best there is. Sometimes, save money at the beginning is not a 'savings' per se but ends up being the cause of a lot more expenses later. Expenses that could have been avoided (or reduced) by spending a little more in the beginning.
Hold on, I'm not saying that you should spend a lot of money right away, but my experience has been so far (and still is) that trying to save every single cent, in many cases meand ending up spending a lot more to obtain a decent result. What happens, in my experience, is that you get something paying as little as you can in the hope that it would be functiona, then discover that wathever it is, is not functional enough to do what you need, so you end up buying something else to have the first thing actually working or to replace it. And you keep on doing until you've spent more money than you had spent buying something more expensive but functional right away.
And after this expensive introduction, we'll talk about $weneedmoney, a (more or less) known company that provided financial services.
These peoples have build a bunch of "digital services" that, based on their idea, should have allowed their customers to do more work in less time and so... Ok, you know it. The problem was that, as per previous experiences, they had decided that the best thing to do (from their point of view) was to spend as little as possible to set up all those '"services".
And one of the bases for this stuff was the hosting service. That had been selected to provide the maximum service with the minimal cost. And then they began to think that maybe, the "maximum service" doesn't need to be that "maximum" after all... Something less than that is even better... Even less than that? Can't we do it for cheaper?
At the point that, one of the principal component for the entire infrastructure, the disks (or what have you nowadays), had been reduced to the minimun required and then replaced by a NAS bought by $weneed itself. A NAS that, as we discovered later, had been bought at a bankrupt auction after a couple of years spent in some warehouse.
With this I mean that the unit wasn't exactly brand-new.
And still with the goal of spend as little as possible, their hosting plan included the total of ONE backup with 30 days retention. Now, I've no idea what kind of data these people were archiving and how, but I've always considered a less-than-daily backup as a not-a-backup. But who am I to question their money sense?
Karma the crocodile came knocking at $weneed's door about a year later they've started the snake oil sales, in a moment when all their "customers" also began to think that maybe, but just maybe, if it shine it doesn't have to be gold and a lot of the services sold (at high price) didn't actually provided a value that justified the amount written on the invoices.
Until, one nice morning, we got a mail that notified us that the storage unit was 80% full.
Ok, (me think), let's see what's the problem. Unfortunately, I couldn't see any problem, because there was no partition in that thing that was 80% full. However, the stupid thing kept insisting that something was full. After a long series of curses and checks, more curses than checks, I found the problem in a well-known bug in the firmware of this thing. A bug that was never fixed by the manufacturer.
Apparently, this thing doesn't reset correctly some internal parameter related to the log partition when the logs are rotated. The result is that, if you ask the controller of this thing, the partition keeps filling up even when it doesn't.
Ok, is time to open a call to the manufacturer of this thing. And then I didn't thought about it a lot until a couple of days later, $weneed, in the person of CL, phone up.
CL - Why do we get this alert that the disk is full?
Me - As I've explained in the ticket that I sent you, it is a bug in the firmware of the storage unit, we've opened a call to the manufacturer and we're waiting for them.
CL - What does "the disk is full" mean?
Me - Is not full, is a bug in the firmware that reports it as full.
CL - But why is it full? Can't we clean it up?
Repeat the 'conversation' at least a dozen times.
Finally, after a few days, a mail arrives from the manufacturer's support. Mail that more or less says:
From our archives it seems that the unix XYZ with serial number HOLYCRAPHOWLONGISIT is End of Life and End of Support since $aboutayearago, as such every support or activity has to be invoiced as "special task". An offer for the job will be sent by our sales department.
I look at it for a bit and then do some more check and... yeah, this "thing", that was already old before, is now completely out of support. I dutifully turn the mail to $weneed, that have bought this thing and as such are the ones that will have to fork the money. Of course they are very very unhappy about forking anything.
And so we begin a very long forth and back between $weneed and the manufacturer of this thing, the base of which is that the root of the problem is a firmware bug that the produced knows about and as such should be covered no matter what's the support status of the device.
Now, I could agree with this, but the reality is that $weneed knew that this thing was old and would've been out-of-support in the short term, but they decided to buy it anyway and base the whole infrastructure on it just to save money.
After about a week, we get the offer. And in reading it I can't help but think if the manufacturer decided to get rid of $weneed as customer or are they kidding. Yes because for an activity that takes probably less than an hour of remote assitance, they ask for about 8000$ plus "expenses" (?). No, you've read it correctly. I asked myself how much would've costed a brand new NAS. And how much for one of a different manufacturer.
But there is nothing to worry about! $Weneed sees and does... We get another mail that $weneed decided to decline the offer (nobody expected anything else) and proceeded to acquire another NAS that will have to replace the old one. The new NAS... Isn't exactly new either. In fact, is maybe older than the original. But the important is to save money right?
So we spend 2 days moving the data from the old one to the older one, and at this point we discover that the new unit is maybe a bit too small... At the point that we cannot copy all the data.
So, result is: one of the unit is on his way to electronic death and the other one is too small.
And as expected, $weneed decide that the best thing to do is... buy a third unit! Again, a used one.
They. Never. Learn. Anything. Fuck.
Davide
13/05/2019 12:55
