Most people that interview with me hate me for it - but in return I've yet to hire someone who didn't do a good job... You first ask a few simple questions so they relax a little - repeat that if they don't know its alright - just say how you would figure it out... Then you move on to questions you're certain they don't know... and ask how they would troubleshoot it. There are always the basics - like ask the user when this first happens, ask exactly what is slow and so on - that shows how they go about a new problem - and in the end, that's what matters. Don't go by how correct their answers are - but how they answer it. If they try to BS their way around it. If they admit they don't know or if they come up with a million different answers. You want the ones that come up with many different things - even if many of them are wrong - it proves that they have the right attitude.
So, now that management has run out of ways to prove that their plans work they will find a new, even cheaper place... good luck with that.
So far I have not come across many Fortune 500s where outsourcing actually worked in the end - that means not just a lower rate but comparable quality. There are plenty of CxOs that announce how much money they saved and all, but if you talk to the techs they almost consistently have another story to tell. For each 100 hours of outsourced work I estimate the average will be about 40 hours of US time to review and fix the programs... And those 40 hours will eat up all the cost savings you had in the original 100 hours. Its sad - but in the end for a million line codebase that has a certain quality, it doesn't really matter where you do it - the cost will be the same... The only ones that have a big advantage there is the russians. No idea why but their quality is usually better than you find anywhere else and the prices are reasonable too.
Before outsourcing, look beyond the hourly rate and consider skills. Then analyze your savings after the project has been in production for a while - and check if your expectations actually came true.
Or lack there off in your case... We have more than 800 ide disks in the systems I maintain. We have a very low failure rate - and if something fails its likely a ibm/hitachi disk.
Anyway, we used to test all drives when they came in... We quickly stopped that and didn't look back. We only have a couple of failures a week (if that) and all our RMA replacements except one came in just fine. WD seems to be the best there - I only got one drive labeled as remanufactured, all other rma drives we got from them were brand new.
They missed one biiiiig issue there... In the US, Europe, Japan and Australia, there are good laws that they can use to come after you... If you move work to India, China or similar, its virtually impossible to get anything from that individual - hence the person has much less worry about doing something illigal...
I don't quite get the point... If all apps have to be signed before install, then you have a point of attack. Intercept communications, fake checksums, attack the OS providers server,... wouldn't be much more secure than anything else.
Wouldn't it make more sense to go back to the live cd concept... You pick everything you need and then make a bootable cd out of that. We did that 10 years ago - was a lot of work but worked great. I'm sure over the years people have written better scripts than the hacks we did back then - but basically you would make a chroot filesystem, put all your apps and do all your testing until stuff worked. Then we created a boot floppy and out of that and the content of the chrooted filesystem, we made a bootable cd. That went into the production server (a high profile site back then that drew tons of attacks) and if the box got hacked, they still couldn't do anything with it.
That way you have 100% security from the point you were looking at - while still maintaining the ability to add/remove/modify things yourself as you need them...
Am I the only one that's confused about the statement? If there is energy in empty space - then that means only one thing - the space by definition isn't empty...
We thought that there was nothing in water - then they found minerals and all other kinds of stuff... we thought there was nothing smaller than an atom - we were proven wrong... So why is everyone so surprised that we found yet another thing that we didn't know existed? Why does it have to conflict with physics? If the particles are so small we just didn't see them before - then they don't have to influence things strongly enough to make a difference to our current physics. Just like when they discovered that the atom wasn't the smallest piece of matter... It opened up a new world - but it didn't relly change the physics that dealt with the higher levels...
Each cable that transports AC is subject to drain by the capacity the parallel lines themselves represent. The closer the wires, the higher the capacity. At about 30 km on a regular high voltage cable, you reach a point where the reactive power drain reaches the maximum power the cable can transport - the cable is saturated without draining a single watt at the end.
DC does not have this issue however then you have all the problems that killed Edison's original DC power distribution in favor of Telsa's AC distribution.
Am I the only one disappointed in Jonathan Schwartz? I met him a few times and while he seemed very knowledgeable and charismatic, he lacked quite a bit in mental flexibility. I doubt it was my comments on the panels I was on, but he pretty much laughed in my face when I made certain comments - and now, two years later the software at sun is treated pretty much in exactly the way I talked about. Now, its Sun's money and whatever - but that (and a few similar incidents with him) left a bad taste in my mouth. He does a good job at hiding it, but he's incredibly arrogant and suffers under the not-invented-here issue that already brought down many companies of substantial size - think Digital.
Sorry to say that - but Sun needs someone who is more open and listens better than Schwartz. He's a good leader, but he certainly lacks in vision and new, revolutionary ideas.
I work in a department that creates provision software for one of the large telcos... As you said, the problem is usually application... OS, DB and such are usually no issue anymore but unfortunately "Enterprise" in the US seems to mean disorganized mess with completely incompetent management - management that rather wants to keep pointless dates that have no meaning in the real workd than doing things right.
We have a few well designed apps and there the answer is pretty much "How big a machine can you buy" - scalability isn't an issue. As I said before, Oracle, DB2 and Informix all scale well enough these days to span 64 or more cpus without issues _IF_ you have your table design right. OS tuning is fairly simple these days. DB tuning takes a little more skill but that's not too hard either. Getting the right hardware set up is next to impossible because of processes. So, no matter what direction you look, it all ends up on the development team's shoulders.
If you asked a little more specific, I'm sure you would get better answers...
Not really - AMD doesn't have dual core laptops out yet and even if you look at just one core, the Intel Core architecture beats anything AMD has right now hands down...
Hey! I also develop a cross platform tool... Its the thing to do - everyone else does it this week!
Seriously, why would I do that? We have all those management tools that exist already and none of them work... Not even monitoring works. The tools we have (fortune 15 company) cost us millions, and none do what we need. We can't even monitor servers running vmware! VMWare servers have been out now for what? 3 years? And the only one that even claims they can do it is HP - and their stuff is still far from painless.
Grid, Utility Computing, server solsolidation efforts and whatever you want to call them - it has never worked... and now we have yet another try... *shrugs* Sorry if I'm not excited...
If you're in a situation where you really have to worry that much about your own people, doesn't that just show that management has failed to provide a good working environment and create loyalty?
The only effect of security is going to be that the few loyal employees you have get pissed and turn against you too. And for anyone who has done only a little bit of hacking, we all know useful security is way too expensive... You'd need to audit virtually everything that's going on on a server and there are only a few government agencies that can efford that much money.
So why not do something more useful with the money? Free coke for employees on tuesdays. Or fix that darn pothole at the entrance of the parking lot. Put a few plants up in the office... That is all money better spent than on some lack luster, process bound security measures...
Is anyone really surprised with that? In a time where a company's stock price can fall by 50% and the execs get 50 million bonus while the employee's pensions are being cancelled, how can you say anything good about management?
And I figured out how to win the next Superbowl - all I have to do is score more points than the opponent...
But seriously - the article is pretty useless... The hard part is not to understand that you have to do something new and excellent - its how to do it. A minor detail the article unfortunately doesn't explain...
Your question is a good one... To my knowledge it is also one that nobody has an answer to. I know three different companies that try this in depth - one of them has several dozen different projects with different managers. And no one has figured out a way to manage a project if communications and cooperation are important.
I know higher management never wants to hear it but projects fair well with offshoring if you either offshore the whole thing (and US management gets out of the way) or if the US based team is strong enough to completely ignore the remote centers and do the whole job... There are whole books out there that talk about this topic - but nothing works well in real world.
You know - its an big old telco... That means stupid rules, red tape and the most ridiculous processes you can imagine... Trust me - anyone who has worked for Bell Atlantic or VZ will be able to tell you that... Time reporting has to be done by friday morning (some orgs even wednesday evening) - yet you are responsible to accurately record your time till saturday.. Yeah - I'm not planning on having any outages on Friday, you know... If you live in a world like that, that statement about Google almost makes sense. Inside the Verizon Reality Distortion Field(tm) of course...
For the outside world its a different story... Its like asking a mail order company for a share of the profits just cause they use your phone lines... Or if you sell your shares in company X you should have to pay part of the profits you made - after all you're using verizon lines...
Anyway - this is just the reaction of an old dinosour that's ready to collapse under its own weight. They throw money at FTTP - something that has a 40 year projected return on investment... Somehow they need to make money. And if they lobby for it and it goes through, they'll have another way to extort money...
So many replies about performance and all the great new features in SATA IO... Too bad no one has mentioned the real issue for high end use... All modern scsi drives allow the HBA to make sure data has actually been written to disk. SATA does not. If you lose power, you lose the data in cache. Worse even in a raid - there the data might be written to two of the data disks but your redundancy disk(s) have not. Now, if a drive fails, you'll restore from parity and boom - you have wrong data that according to all drives, controller and filesystem status is good. That's a real bad thing(tm).
SCSI controllers with battery buffered cache will take care of that - Unless all drives have written the data to disk, the line is not released in the controller cache and when power comes back on, it can be written to the drive. That will not keep you from having to check consistency but it will prevent you from seeing corrupt data as good.
First make sure everyone knows the rating schedule. 10% of your organisation are Leaders, 70% are performing at an acceptable level and 20% are not good enough. Doesn't matter if you hire the best of the best or not - its always going to be like that. After all, you're management and know better:-)
Next make sure that you motivate your troups. Constant layoffs are a good way. Make sure though that you send out emails stating that the layoffs are because of budget constraints and at the same time you give yourself a nice raise and a several million dollar bonus.
After that, hold a great big meeting. When there, make sure you include a phrase like "To perform well, we need to attrakt top tallent. We can only do that in India" - this will impress your folks that worked their asses off for years.
Finally, make sure that the people that are still motivated get nothing done. This is best achived by putting so many stupid processes in place that noone knows what to do. That will frustrate them and they will eventually stop trying..
Once you followed through that your job is real easy - every single person that has not quit yet, is so bad that they can't leave your shop. You can just all rate them as underperformers.
But just in case your question was meant seriously here is a less sarcastic answer. Remember, Management doesn't produce anything. Management is useless overhead unless it is put to work _for_ the people that actually produce the product the customer is asking for. So ask your staff what they would need from their management to be more productive. If they can answer that question, then they care and try to do a good job. If they don't have an answer, they don't care.
Who cares about performance? If you want to do something useful, then test reliability... Pull the power on the server. Have a failure on a drive, see if you can rescue any data. Corrupt a sector by overwriting it with 0s and see what the engine does.
Performance you can (almost) always go to a bigger box - if your whole engine goes down and can't recover because of a few bad sectors or insufficient logging or such, you're screwed.
I think that while many of those business cases have a good idea, the real question is how to best sell your idea. After all the greatest idea in the world will get laughed at when if comes from a guy that looks like Urkel... So what did those 10 buisness cases have that made them great? How were they presented? I bet those cases aren't any better than a million others that failed just because they were not presented adequately.
The key is to show the value of education... I grew up in Austria and I never bothered studying my English vocabulary. In fact, once I argued with my teacher that I have no use for it - so why would I learn it... 10 years later, I surely do know the value of it - after all I'm now happily living in Florida.
My gf's daughter had a similar experience... She lived for a few years in Plant City, FL - redneck and pedophile heaven. Now that she lives in a more civilized part of the state and people often have a hard time following her grammer and pronounciation she's starting to learn the value of it. One day she got into a fight with one of her best friends over a misunderstanding caused by her language. Ever since she's trying really hard and is much easier to deal with when it comes to "making her study"..
So in the end - find a way to show the student why they needing and how they benefit and the studying is suddenly no issue at all anymore...
Windows on itanium is a joke... What software are you going to get running well there? We tried it and 80% of the software we needed to certify a new OS wasn't there.
HP-UX is better off but still - if you have any legacy software at all in your system you're screwed.
Linux is doing alright - but if you use a Itanium box running Linux and pit it against new xeon with the same number of CPUs, the Itanium looks like a dog...
Most business apps are integer processing - itanium doesn't look that great in the int benchmarks...
I'm frankly just tired of hearing about it... Since 7 years we hear that itanium is going to be the future and all - hasn't happened yet and I doubt it ever will at the pace its moving. Why port to a platform that already feels dead before it even took off?
You can't compare a Xeon and an Itanium box by the per cpWe already support 5 different platforms - why would I want to add a 6th one if the performance gains are going to be pretty meger...
Well, first off - invention is more difficult these days... invent a new computer chip? Oh yeah - once you got a billion dollar fab that's easy...
Second its out patent system and its ridiculous enforcement. In a system where a case like the one with SCO can go on for so long without them showing any prove who can expect you to invent stuff?
Slashdot Mirror
Most people that interview with me hate me for it - but in return I've yet to hire someone who didn't do a good job...
You first ask a few simple questions so they relax a little - repeat that if they don't know its alright - just say how you would figure it out... Then you move on to questions you're certain they don't know... and ask how they would troubleshoot it. There are always the basics - like ask the user when this first happens, ask exactly what is slow and so on - that shows how they go about a new problem - and in the end, that's what matters. Don't go by how correct their answers are - but how they answer it. If they try to BS their way around it. If they admit they don't know or if they come up with a million different answers. You want the ones that come up with many different things - even if many of them are wrong - it proves that they have the right attitude.
Peter.
So, now that management has run out of ways to prove that their plans work they will find a new, even cheaper place... good luck with that.
So far I have not come across many Fortune 500s where outsourcing actually worked in the end - that means not just a lower rate but comparable quality. There are plenty of CxOs that announce how much money they saved and all, but if you talk to the techs they almost consistently have another story to tell. For each 100 hours of outsourced work I estimate the average will be about 40 hours of US time to review and fix the programs... And those 40 hours will eat up all the cost savings you had in the original 100 hours. Its sad - but in the end for a million line codebase that has a certain quality, it doesn't really matter where you do it - the cost will be the same... The only ones that have a big advantage there is the russians. No idea why but their quality is usually better than you find anywhere else and the prices are reasonable too.
Before outsourcing, look beyond the hourly rate and consider skills. Then analyze your savings after the project has been in production for a while - and check if your expectations actually came true.
Peter.
Or lack there off in your case... We have more than 800 ide disks in the systems I maintain. We have a very low failure rate - and if something fails its likely a ibm/hitachi disk.
Anyway, we used to test all drives when they came in... We quickly stopped that and didn't look back. We only have a couple of failures a week (if that) and all our RMA replacements except one came in just fine.
WD seems to be the best there - I only got one drive labeled as remanufactured, all other rma drives we got from them were brand new.
Peter.
They missed one biiiiig issue there... In the US, Europe, Japan and Australia, there are good laws that they can use to come after you... If you move work to India, China or similar, its virtually impossible to get anything from that individual - hence the person has much less worry about doing something illigal...
Peter.
I don't quite get the point... If all apps have to be signed before install, then you have a point of attack. Intercept communications, fake checksums, attack the OS providers server, ... wouldn't be much more secure than anything else.
Wouldn't it make more sense to go back to the live cd concept... You pick everything you need and then make a bootable cd out of that. We did that 10 years ago - was a lot of work but worked great. I'm sure over the years people have written better scripts than the hacks we did back then - but basically you would make a chroot filesystem, put all your apps and do all your testing until stuff worked. Then we created a boot floppy and out of that and the content of the chrooted filesystem, we made a bootable cd. That went into the production server (a high profile site back then that drew tons of attacks) and if the box got hacked, they still couldn't do anything with it.
That way you have 100% security from the point you were looking at - while still maintaining the ability to add/remove/modify things yourself as you need them...
Peter.
Am I the only one that's confused about the statement? If there is energy in empty space - then that means only one thing - the space by definition isn't empty...
... It opened up a new world - but it didn't relly change the physics that dealt with the higher levels...
We thought that there was nothing in water - then they found minerals and all other kinds of stuff... we thought there was nothing smaller than an atom - we were proven wrong... So why is everyone so surprised that we found yet another thing that we didn't know existed? Why does it have to conflict with physics? If the particles are so small we just didn't see them before - then they don't have to influence things strongly enough to make a difference to our current physics. Just like when they discovered that the atom wasn't the smallest piece of matter
Peter.
Each cable that transports AC is subject to drain by the capacity the parallel lines themselves represent. The closer the wires, the higher the capacity. At about 30 km on a regular high voltage cable, you reach a point where the reactive power drain reaches the maximum power the cable can transport - the cable is saturated without draining a single watt at the end.
DC does not have this issue however then you have all the problems that killed Edison's original DC power distribution in favor of Telsa's AC distribution.
Peter.
Am I the only one disappointed in Jonathan Schwartz? I met him a few times and while he seemed very knowledgeable and charismatic, he lacked quite a bit in mental flexibility. I doubt it was my comments on the panels I was on, but he pretty much laughed in my face when I made certain comments - and now, two years later the software at sun is treated pretty much in exactly the way I talked about. Now, its Sun's money and whatever - but that (and a few similar incidents with him) left a bad taste in my mouth. He does a good job at hiding it, but he's incredibly arrogant and suffers under the not-invented-here issue that already brought down many companies of substantial size - think Digital.
Sorry to say that - but Sun needs someone who is more open and listens better than Schwartz. He's a good leader, but he certainly lacks in vision and new, revolutionary ideas.
Peter.
I work in a department that creates provision software for one of the large telcos... As you said, the problem is usually application... OS, DB and such are usually no issue anymore but unfortunately "Enterprise" in the US seems to mean disorganized mess with completely incompetent management - management that rather wants to keep pointless dates that have no meaning in the real workd than doing things right.
We have a few well designed apps and there the answer is pretty much "How big a machine can you buy" - scalability isn't an issue. As I said before, Oracle, DB2 and Informix all scale well enough these days to span 64 or more cpus without issues _IF_ you have your table design right. OS tuning is fairly simple these days. DB tuning takes a little more skill but that's not too hard either. Getting the right hardware set up is next to impossible because of processes. So, no matter what direction you look, it all ends up on the development team's shoulders.
If you asked a little more specific, I'm sure you would get better answers...
Peter.
Not really - AMD doesn't have dual core laptops out yet and even if you look at just one core, the Intel Core architecture beats anything AMD has right now hands down...
Peter.
Hey! I also develop a cross platform tool... Its the thing to do - everyone else does it this week!
Seriously, why would I do that? We have all those management tools that exist already and none of them work... Not even monitoring works. The tools we have (fortune 15 company) cost us millions, and none do what we need. We can't even monitor servers running vmware! VMWare servers have been out now for what? 3 years? And the only one that even claims they can do it is HP - and their stuff is still far from painless.
Grid, Utility Computing, server solsolidation efforts and whatever you want to call them - it has never worked... and now we have yet another try... *shrugs* Sorry if I'm not excited...
Peter.
If you're in a situation where you really have to worry that much about your own people, doesn't that just show that management has failed to provide a good working environment and create loyalty?
The only effect of security is going to be that the few loyal employees you have get pissed and turn against you too. And for anyone who has done only a little bit of hacking, we all know useful security is way too expensive... You'd need to audit virtually everything that's going on on a server and there are only a few government agencies that can efford that much money.
So why not do something more useful with the money? Free coke for employees on tuesdays. Or fix that darn pothole at the entrance of the parking lot. Put a few plants up in the office... That is all money better spent than on some lack luster, process bound security measures...
Peter.
Is anyone really surprised with that? In a time where a company's stock price can fall by 50% and the execs get 50 million bonus while the employee's pensions are being cancelled, how can you say anything good about management?
Peter.
"It should be an excellent game"
And I figured out how to win the next Superbowl - all I have to do is score more points than the opponent...
But seriously - the article is pretty useless... The hard part is not to understand that you have to do something new and excellent - its how to do it. A minor detail the article unfortunately doesn't explain...
Peter.
Your question is a good one... To my knowledge it is also one that nobody has an answer to. I know three different companies that try this in depth - one of them has several dozen different projects with different managers. And no one has figured out a way to manage a project if communications and cooperation are important.
I know higher management never wants to hear it but projects fair well with offshoring if you either offshore the whole thing (and US management gets out of the way) or if the US based team is strong enough to completely ignore the remote centers and do the whole job... There are whole books out there that talk about this topic - but nothing works well in real world.
Peter.
You know - its an big old telco... That means stupid rules, red tape and the most ridiculous processes you can imagine...
Trust me - anyone who has worked for Bell Atlantic or VZ will be able to tell you that... Time reporting has to be done by friday morning (some orgs even wednesday evening) - yet you are responsible to accurately record your time till saturday.. Yeah - I'm not planning on having any outages on Friday, you know...
If you live in a world like that, that statement about Google almost makes sense. Inside the Verizon Reality Distortion Field(tm) of course...
For the outside world its a different story... Its like asking a mail order company for a share of the profits just cause they use your phone lines... Or if you sell your shares in company X you should have to pay part of the profits you made - after all you're using verizon lines...
Anyway - this is just the reaction of an old dinosour that's ready to collapse under its own weight. They throw money at FTTP - something that has a 40 year projected return on investment... Somehow they need to make money. And if they lobby for it and it goes through, they'll have another way to extort money...
Peter.
So many replies about performance and all the great new features in SATA IO... Too bad no one has mentioned the real issue for high end use... All modern scsi drives allow the HBA to make sure data has actually been written to disk. SATA does not. If you lose power, you lose the data in cache. Worse even in a raid - there the data might be written to two of the data disks but your redundancy disk(s) have not. Now, if a drive fails, you'll restore from parity and boom - you have wrong data that according to all drives, controller and filesystem status is good. That's a real bad thing(tm).
SCSI controllers with battery buffered cache will take care of that - Unless all drives have written the data to disk, the line is not released in the controller cache and when power comes back on, it can be written to the drive. That will not keep you from having to check consistency but it will prevent you from seeing corrupt data as good.
Peter.
Once you followed through that your job is real easy - every single person that has not quit yet, is so bad that they can't leave your shop. You can just all rate them as underperformers.
But just in case your question was meant seriously here is a less sarcastic answer. Remember, Management doesn't produce anything. Management is useless overhead unless it is put to work _for_ the people that actually produce the product the customer is asking for. So ask your staff what they would need from their management to be more productive. If they can answer that question, then they care and try to do a good job. If they don't have an answer, they don't care.
Peter.
Who cares about performance? If you want to do something useful, then test reliability... Pull the power on the server. Have a failure on a drive, see if you can rescue any data. Corrupt a sector by overwriting it with 0s and see what the engine does.
Performance you can (almost) always go to a bigger box - if your whole engine goes down and can't recover because of a few bad sectors or insufficient logging or such, you're screwed.
Peter.
I think that while many of those business cases have a good idea, the real question is how to best sell your idea. After all the greatest idea in the world will get laughed at when if comes from a guy that looks like Urkel... So what did those 10 buisness cases have that made them great? How were they presented? I bet those cases aren't any better than a million others that failed just because they were not presented adequately.
Peter.
The key is to show the value of education... I grew up in Austria and I never bothered studying my English vocabulary. In fact, once I argued with my teacher that I have no use for it - so why would I learn it... 10 years later, I surely do know the value of it - after all I'm now happily living in Florida.
My gf's daughter had a similar experience... She lived for a few years in Plant City, FL - redneck and pedophile heaven. Now that she lives in a more civilized part of the state and people often have a hard time following her grammer and pronounciation she's starting to learn the value of it. One day she got into a fight with one of her best friends over a misunderstanding caused by her language. Ever since she's trying really hard and is much easier to deal with when it comes to "making her study"..
So in the end - find a way to show the student why they needing and how they benefit and the studying is suddenly no issue at all anymore...
Peter.
Peter.
Well, first off - invention is more difficult these days... invent a new computer chip? Oh yeah - once you got a billion dollar fab that's easy...
Second its out patent system and its ridiculous enforcement. In a system where a case like the one with SCO can go on for so long without them showing any prove who can expect you to invent stuff?
Peter.
At least now I know where my tax money went! Termite tents! That's at least more believable than that $15,000 toilet seat and $20,000 hammer :-)
Peter.
Wonder what Linus has to say about the G5 he's using these days to check out if POWER is really such a great architecture...
Peter.