"BearItAll" <spam@xxxxxxxxxxxxx> wrote in message
news:1184057059.72833.0@xxxxxxxxxxxxxxxxxxxxxxx
>
> Not wanting to be anti-MS-IT-Staff, but I think everyone in this buisiness
> knows already that on average UNIX/Linux IT folk tend to be more switched
> on to IT tasks.
>
> We don't expect or go looking for ready made solutions to all problems, it
> is nice when a distro does do some of the grafting for you, for example
> SLES with Yast. But still, I think that most of us are happy to get our
> hands dirty. In fact I would go further and say that is essential to the
> sort of minds that are attracted to become UNIX/Linux managers, otherwise
> we would find it dull.
>
> But having said that I don't know of any UNIX/Linux IT person who doesn't
> get great pride out of the fact that they have very little or no down
> time.
>
> Some of that comes from the IT person foreseeing problems. About 98% of
> all
> server problems give a warning before they are a risk. Not necessarily a
> direct log entry, but using the standard 'Know how your system looks and
> behaves when it is good, then you will see when it is going bad long
> before
> the problem can bother the users'.
>
> So each day I start with a skip around my servers, same pattern each day,
>
> Skip through the logs looking for oddities. Not all problems are
> direct,
> but you get used to pattern searching, things that are wrong stand out.
>
> df -af : Again pattern searching, I know what it should look like on
> each server and I know how quickly it changes. A volume that I have
> arrange
> to never change unless I do it, 1% increase or decrease and I'm onto it.
>
> netstat .... and verious other regular daily checks. The whole thing is
> less
> than 15 minutes.
>
> But it means I get to treat and diagnose any problem very early. That in
> turn has given me a down time record of 8 hours in 4 years and two
> lightening strikes account for most of that.
>
> Now, I do have to deal with quite a few Windows IT staff. Some are very
> clever, some are absolute crap. But there is one thing that they all have
> in common. They all wait for the problem/fault/errors to arise before they
> go looking for them.
>
> That to me is the main difference between the Windows IT mentallity and
> the
> UNIX/Linux mentality. It is also a major part of the reason why MS Win
> servers have such a bad reputation in the general office environment.
>
> I have no doubt at all that many of the Windows server problems could be
> foreseen in the same way as Linux server problems, but they simply do not
> do it. Even as I lecture them to try to get them motivated into wanting
> fantastic uptime records, I know by the look they give that 'This is not
> the Windows way to do things'.
>
> Downtime costs money, Windows gets lots of down time, Linux gets very
> little, it doesn't get very little because our hardware is better, it gets
> very little because it attracts the sort of staff that are willing to put
> in the effort early in order to save them the sleepless nights at a later
> date.
>
> Do you know any Windows IT people? Well ask them a question, ask them how
> confident they are that they can recover their full system from the
> backup.
> Offer them one hundred pounds to perform the full recovery 'now', so that
> they down't get a chance to doctor the backup, you want them to recover
> their normal daily backup tape. I bet you both socks they wont take that
> bet. I didn't offer £100 (I'm a poor IT person) but I have offered a slap
> up indian meal many times, but never had to buy it.
>
> How can they do that? What sort of mentality does it take for an IT person
> to sit there day after day, not knowing if their backup tapes can be used
> to bring the system and data back to how it was at the time of last nights
> backup? Could you live like that, I know I couldn't. I have had many a
> sleepless night over IT problems over the years, but I will never be
> caught
> without my safetynet, and I am certain that the vast majority of
> UNIX/Linux
> IT folk are the same.
>
> How about Budgets then.
>
> My budget here, when I arrived they had a massive IT budget, it was sort
> of
> assumed that IT costs a lot and there's no choice but to pay it. Within a
> year I had cut the IT budget to less than a quarter and improved uptime,
> and put in place a very secure set of safetynets which are tested
> regularly. I can sleep at night very well.
>
> So this nonsense about Windows IT being cheaper than Linux IT, is pure
> crap.
> Yes the staff will cost more in salery, but the savings from uptime and
> hardware costs make it an excellent investment.
>
> But all of this uptime generates a problem unique to UNIX/Linux. There is
> a
> mentality in general management that if the servers and computers are not
> going wrong then the IT staff are not doing anything. They have an idea
> that we only work when they is a problem on the system.
>
> I think I have convinced mine over many years, but I could understand if
> others around the world are having more trouble convincing their
> management
> that the fact that they have a lot of uptime is proof that the IT folk are
> doing their job, downtime is a failure of the IT task. This again comes
> from a Windows-like mentality, where the system goes down, the IT person
> turns up and fixes it, then they proclaim him/her a hero for fixing the
> system. I tell them though that he/she isn't a hero, he/she failed to keep
> their network in proper working order, some problems are inevitable, but
> if
> you see your IT staff firefighting a lot, they are doing a crap job of
> looking after your system. (I'm not very popular at IT staff parties).
>
>
I quite agree with all that, and in addition, I also found (wearing my
Systems Accountant Hat) that many senior managers are actually afraid of the
IT dept - my philosophy being that the IT dept is there to provide a service
to the organisation, not the other way around. As an example - I worked some
years ago as Group management Accountant to an IT PLC. We used JD Edwards
accounting, which was not very good at reporting with data integrity. So we,
the Group Finance dept decided, after some research, that we needed to
implement Hyperion, which at that time was one of the industry-standard
third-party reporting tools. So we bought it. the IT dept threw a hissy fit
because we hadn't "consulted" them and they hadn't "tested" it. Industry
Standard? As far as I know it is still sitting on the shelf, some 16 years
later. Why? because senior management didn't tell the IT dept to get a grip,
this is what we need, implement it.
|
|
