S
Appliance testing - none of you would do this would you ?
- Thread starter SimonH2
- Start date
- Status
Sponsored Links
I never said it wasn't, so why are you asking ? Yet again, you are phrasing your questions so as to imply I said something I didn't. As you well know, and have even admitted to, what I said was that not all virtualisation configurations can do live migration. You even went as far as to say "... and without a common storage and network infrastructure shared by source and target then live migration won't work"
Tell you what, since it was YOU that introduced virtualisation to the thread, how about you answer one simple question.
Just one simple question.
A really simple, yes or no, question.
Given what has been discussed about virtualisation, and your previous admission that not everything that fits the term virtualisation can do live migration, can it be an untrue statement to say "So virtualise them all, then you can move the VMs and empty the physical server." (the emphasis on "can" is mine).
(I've sent out for popcorn
)Ooh, so this is why BAS mentioned toasters
As an old timer who has designed a lot of hardware and software I am amazed and concerned that just changing software can make hardware overheat to the extent that systems start to shut down. If the software "update" affected the hardware monitoring systems then OK but if it was changes to applications ( progams ) dealing only with data etc then that system is too critical in my ( old fashioned ) opinion.
Sponsored Links
They talk about firmware updates - so that could have borked and left them running flat out (CPUs on full chat, fans on full). Just having CPUs on full chat all teh time could be enough, but I suspect that having all the fans running full speed would upset the airflow in the datacentre no end.
Reminds me of the first XServe I worked on, when they were brand new (we had to wait for the RAID card to become available). A config problem on the RAID card left it unable to boot - but you could hear the fans from the other end of the building
Reminds me of the first XServe I worked on, when they were brand new (we had to wait for the RAID card to become available). A config problem on the RAID card left it unable to boot - but you could hear the fans from the other end of the building
Oh goodie, a grown-up discussion we can get into without BAS' ego getting in the way (I hope!).
I've seen what both Bernard and Simon mention - both an application update that causes CPU or RAM thrashing (ironically, I seem to recall that an ESX update was one such cause) or a BIOS update that upsets the fans.
Our aircon units now respond to pings and send snmp traps so we can tell if they are having a problem (snmp) or fail entirely (ping).
Before we had such exciting technology, we would rely on the climate monitors but, ironically enough the first hint of trouble was when the sound levels increased. That was a sign that *all* the servers' fans were running at full tilt in a vain attempt to lower the temperature.
We knew we were in the crap if we someone said "that's odd - just had an alert that the sound level in the server room has increased..."
Five minutes later we'd get all manner of over-temperature alerts
I've seen what both Bernard and Simon mention - both an application update that causes CPU or RAM thrashing (ironically, I seem to recall that an ESX update was one such cause) or a BIOS update that upsets the fans.
Our aircon units now respond to pings and send snmp traps so we can tell if they are having a problem (snmp) or fail entirely (ping).
Before we had such exciting technology, we would rely on the climate monitors but, ironically enough the first hint of trouble was when the sound levels increased. That was a sign that *all* the servers' fans were running at full tilt in a vain attempt to lower the temperature.
We knew we were in the crap if we someone said "that's odd - just had an alert that the sound level in the server room has increased..."
Five minutes later we'd get all manner of over-temperature alerts
I know what you mean, sometimes "the server room sounds warm" is heard in our place. As you say, you can hear the effect varying temperature has on the fans.
Incidentally, I've even been able to correlate it with power consumption. With a load of around 7kW, I've been able to pick out around 200 to 300W of variation that correlates with temperature. I did consider that it's variation could be due to changing load, but the variation is significantly less on cool days.
We don't use chillers, only air cooling.
EDIT: And as an aside, I was told that experienced exchange engineers (back in the days of electromechanical telephone exchanges) could tell the general state and 'health' of the exchange just by it's sound.
Incidentally, I've even been able to correlate it with power consumption. With a load of around 7kW, I've been able to pick out around 200 to 300W of variation that correlates with temperature. I did consider that it's variation could be due to changing load, but the variation is significantly less on cool days.
We don't use chillers, only air cooling.
EDIT: And as an aside, I was told that experienced exchange engineers (back in the days of electromechanical telephone exchanges) could tell the general state and 'health' of the exchange just by it's sound.
Ah... the old days...
I started, way back when BAS was still in his 60s, repairing DEC teletypwriters that received data on either 300/300 baud or 1200/75 baud lines. You could tell when the hosts (typically PDP-11) were in difficulty because the noise from the teletypwriters slowed down when the PDP couldn't even pump data out at 300baud
I started, way back when BAS was still in his 60s, repairing DEC teletypwriters that received data on either 300/300 baud or 1200/75 baud lines. You could tell when the hosts (typically PDP-11) were in difficulty because the noise from the teletypwriters slowed down when the PDP couldn't even pump data out at 300baud
S
SaladFingers
Why?
WHY?
Because of the way you've carried on ever since I wrote this:
NO I AM NOT.
I just wanted you to make it crystal clear that there was absolutely nothing incorrect about what I wrote. Implying certain prerequisites does not make it wrong, even if the prerequisites are not explicitly mentioned.
And yet despite it being correct, you immediately launched into a sustained campaign to say I was wrong to write it, that I didn't know what I was talking about etc.
No, they can't.
So - how many analogies of your ridiculous criticisms shall I put here?
ME: in response to a comment that a boat to New York takes a long time, "Go by plane".
YOU: because I did not explicitly say more, "Not all planes have the range to get to New York. And you'd need the right sort of luggage. And you can't keep your Swiss Army knife in your pocket. You didn't mention any of those things, so you have only a politician's understanding of aviation, or worse still you're a consultant who's only been reading the airline's glossy brochures".
If you would like to persist, I can come up with more.
So?
Your point is?
When Eric wrote what he did, did I reply "So virtualise them all but don't bother with a common storage and network infrastructure shared by source and target, then you can move the VMs and empty the physical server"?
No.
That statement cannot be untrue because it is not sufficiently qualified to define an environment where VM migration would not work.
It can only be interpreted by someone as being untrue if they deliberately choose to add assumptions to it which would make it untrue.
It is no more untrue than "You can fly to New York" or "You can go by car" are untrue.
Or "You can use a toaster to make toast".
If you attempt those things and fail because they won't let you check in a black binbag as luggage, or there are no roads where you're going, or you have no electricity that does not make the statements untrue.
Is it your position that no comments or observations etc should ever be made on this forum without explicitly and exhaustively listing every prerequisite that would need to be fulfilled?
So, that would be a yes then - the statement can be untrue.
It's a statement you made, I have not removed any qualifications you included in it.
Therefore, if you are saying that it needs qualifying to be true, then it can be untrue if not qualified.
S
SaladFingers
I like this. Bas has been rightfully challenged and is doing his utmost best at saving face.
Says the man demonstrating a politicians level of knowledge of technical matters
DIYnot Local
Staff member
If you need to find a tradesperson to get your job done, please try our local search below, or if you are doing it yourself you can find suppliers local to you.
Select the supplier or trade you require, enter your location to begin your search.
Are you a trade or supplier? You can create your listing free at DIYnot Local
- Status
Sponsored Links
