Acutlay if you ever looked at how DSL works smalls towns take a one time cost to increase bandwith. DSL works as follows:
Home "Modem" takes Ethernet frames stips off the ethernet bits and adds ATM headers at layer 2 and throws it onto the actual DSL line.
The DSLAM general owned by the incumbant telco taks the DSL signaling and puts it into the ATM cloud.
The allready exisitng ATM out of town infrastructure they backhaul this to whever the DSL providers decide to hook up and via whatever sized pipe the provider wants. Because it's ATM if it's a garenteed speed they have to have the pipe to fit it in. The ATM cloud is the only parts the telco has to deal with expanding it's realy one time costs of getting faster line cards as they own the backhauls anyway to support more bandwith. Technicaly because of the way ATM works there cant be bandwith issues in the bachhaul portion because ATM uses two basic types of PVC there is CBR thats fixed rate and EVERYTHING reserves the space for it or VBR that is as allowed. Most DSL I have seen gets provisioned at VBR with a fixed max and min speed all the min speeds added up have to be smaller than the all the connection from point a to b have room for or the PVC wont come up and will never work at all.
Small telcos general have the issue of no close points to backhaul to cheap. If I get a line into New York or most major cities I can get bandwith for $10 a megabit and up with good bandwith at under a hundred. But you have to get the data there first to a rural ISP that might be a long backhaul from some small city thats outragious at the small end.
OK I will have to disagree with the fee structure so far at least in Connecticut I work with several small and midsized ISP's (regional) the DSL fee structure as it stands is in the telcos favor. First off there is a fairly steep ATM access price thats equivilent to a leased line and then some generaly in the neighborhood or 30-45k for an OC3 ATM monthly with a good 100k install fee. Then they want 35 per line plus an install fee. General hassels getting service and support that you see with there own service but not to the same extent. Now on top of this they are offing a 29 a month contract for a year agreement with a waived install. Now if this wasent bad enough they have 11 a month in nebulus DSL access fees charged directly to the users bill. All tole they are making 46 bucks a month per DSL line from a competitor while selling theres for 40. This is clearly a monopoly abusing it's power by subsidising it's DSL and internet server that runs at a loss via revenues from it's traditional services. Funny thing with all of this is if I orded a dry pair and didn't have to use there DSLAM's I could provision the line for 7 a month and the costs of supporting a DSLAM in the switch house or someplace very near by it possibly on the POLE outside (Space rented from the Eletric company pretty inexpensive actualy). I could also backhaul via my own circuts and skip the nebulus ATM cloud and those associated provisioning costs (Doesnt anybody relize ATM is dying a death it deserves it's a dead end technicaly, it's easier to throw more bandwith at the problem than ration everything out cheaper as well)
Having the local government run a copper last mile plant might be a good idea. If you ever have used low density copper it's realy not that hard to maintain unless somehting falls on it you might not deal with it for years. A hybrid fiber and copper plant might be a better idea it's not that hard to convert and or split out to copper. Copper is still just a lot more useable that fiber into the home and getting a few thousand feet of fiber put in it's a lot cheaper and easier that miles worth if you realy want fiber. A side note on fiber it's possible to do it on the pole but cant we finialy get away from poles and stay burring things where they belong?
Whats wrong with a router doing NTP all the routers I configure use it and most of them serve it up as well. This was the same argument people made about not puting DHCP on the router again something thats very usefull to have in place. Yes it's a trade off vs centralized management but in the case of an NTP server what do you manage centraly?
It was bad form for Netgear to use somebody elses NTP server. It would have been more apropriate to have it use the DHCP flag (for those cable modems) for NTP server and use that one if avalible otherwise use some DNS name from netgear.
Ah but she could make an argument that she might be opening herself up to criminal charges by having the information disclosed. You cant use a civil proceding to get around due proccess it what could be a criminal matter or due process is worthless.
I beleive you want an Nforce2 chipset it's not socketed but todays new PC owner dosent change out there video card. But otherwise slots are usefull for neatness sake if nothing else and high speed busses. The fastest external connector is firewire 800 at 100MB a sec half duplex (Not entirly sure ont hat bit) It cant deal with GigE speeds. PCI at base does 132MB a sec and can clock much faster and wider currently to around 1024MB a sec thats 4 GigE adapters running full out to get buss saturation. And most systems that implement PCI-X run multiple buses of it for more potential speed. Firewire 800 is porbably fast enough for nearly everything a home users needs to do except gigabit networking but dosent scale to servers.
Re:Origins of the Internet - no power, no work ?
on
Network Blackout
·
· Score: 2, Informative
Define work? With current routing topologies you take down all the tier ones and your not getting out of the USA and will have trouble getting much farther than that. Contract wise the tier ones have been applying a lot of presure on the tier 2 guys not to advertise interconnects and often have good reasons not to. Add to this the massive ammounts of long haul centralization take out a few MAE points and things would be bad VPN's are replaceing the long haul circut and as it gets nastier and naster out there firewalls are the norm blocking trafic through corps private backbones and satalite links.
So does this mean they can send adds for the cheapest booze in town? Or for eddies Cardboard box emporium? If your going to spam somebody it's generaly commercial and thats the one group thats not interested in the indigent.
As a hiring manager PHD's arent particulary atractive in and of themselves. With the poor state of pratical education persoanly I would rather hire a person with more experience than an over educated one. The only roles that generaly seem to be looking for PHD's are CTO and senior research positions and either of those again want to see a lot of expereince.
OK I dont know where you are or even if your in the USA. But lets not let that stop me making the usual american twist on things. First off in a lot of states even eletrical engeineers arent allowed to work on commercial or industrial wiring it has to be a licened eletritian otherwise a nice building inspector can come in and close the building period end of sentance yes the nice sherif will escort you out now till it's fixed and up to code. Now with that all said rewiring a mess like you describe is anywhere from easy to very complex depending on your UPS and gen set needs. From your description of the main fuses (those nice 1liter bottle looking things) you are running a decent ammount of current the perfered method would be to get a new pannel mount it get your eletric company out to move over the mains while you move over all the circuts. If at all possible oversize the pannel and ask for an increase in incoming amps they never hurt. If you realy need minimal downtime try and get a new main service attached it to the new pannel and move over circut but circut to the live new pannel it's not that dangerious just stay slow and pedantic.
So your saying your compnay had the usual poor design of an M&M hard on the outside soft on the inside. Realy at least they should require corperate standards for laptop firewalling and virus scanning at minimum.
The point is to allow other documenters to make the more human readable source. The definitive technical documentation either has to come first or the origional expert aka the program team needs to make the human readable for dummy's type then write the hard core technical documentation and thats backwards. Your asking probably the wrong person or group to write the dummed down documantation when what should happen is to have an outside interpeter write the dummed down version after working through the expert material.
I mean that first tier documentation should be exceedingly technical this is what the programmer the programming team or possibly the tech writter assigned to such groupings should be writting it. Programmers dont write good manuals in general but they do need to cover the specifics in detail these things are often very cryptic to everybody that hasent been around the particular piece of technology before. But this is the level of documentation required for others to be able to realy use the application.
My general issue with for dummy's books is lack of technical detail they are not clear enough to limitations. Take for example two simple lines talking about remote systems connected to a server application here is the dummy's version at best:
The server can support hundred of users on a modest server.
A better version:
The server utilizes 1 48KB desciptor bound to OS memory block size per active session. The session unique identifer is arbitrarily a unsigned int as defined on the local system.
One just gives you no real information the other gives detained sizing requirements as to the session overhead and max count. Generaly you dont read technical manuals just as you dont ready encyclopedia's (not a great example) For dummy's books are arguably educational program documention should be referance.
I dont hate users I dont like users I generaly dont deal with users thats what help desk is for. If by users you mean System Admins I deal with them often mostly by fixing there issues.
Oh dont worry I make plenty of money working with windows as once people figure out the paper MSCE's cant do what is required they hire a compotent consultant.:)
BTW I like the right tool for the problem beleive it or not clusters of Linux boxes dont work for any problem yet.
Yes documentation should be the expert on the topic written for somebody with a background in real engineering (your average MSCE dosent count) let the howto's and the for dummys books deal with spoon feeding cookbooks to end users if your having the authoritive person on the subject write documentation aka the programming team write the most technical documentation you should ever need without having to do redo code yourself.
I say this because there are to many porly documented applications out there. Documentation to often is looked at by the marketing department and dumbed down so nobody might get scared of it. If you have ever looked at the home service manual for a Saturn (the $500 one thats an option) that nearly would allow you to machine replacment parts thats documentation. Want something easy to read with pretty pictures get a for dummy's book aka the dumbed down book from somebody that read and understood most of the documentation.
Why bother with wastefull wind generation plants (they waste space and generate noise) Fuel cells work can be powered easily off biomass (aka our waste) and with a unit in every house producing 5kw we would have a good deal of excess capacity while keeping generation local the unit sits on a pad outside and is the size of a refrigerator. Realy is you use a pump to compress the natural gas to a storage tank you have a decent backup supply.
For the more power consious attach a battery farm in the middle (ok the eco people dont like these to much but they are reliable) with a larger inverter it could handle larger short peak loads. You still have a power grid to share power from house to house again filling individual peaks (eletric ovens on clean eletric water heaters) Obvious those 50s all eletric houses would need bigger units (or to get a life and forget about eletric heating with the current trends away from nucular power)
OK having worked at a company that everybody had nextel phones they are a great time saver with PTT. You just chirp in as a manager and ask your questions often from meetings if something came along that you werent sure about without having todrag the person there. The big thing is most home users arent responcible enough to turn off the PTT speaker when it's not usefull like in the movies. Add to that that PTT minutes were unlimited on a lot of plans meant a lot of people spent to much time instead of making a phone call. Persoanly Nextel's service is a lot better than Verison simply because it's the only company that realy was pushing to corprate users. Go into a 3rd story underground datacenter from most ISP's that didn't have an affiliated cell phone division and Nextels would work do to repeaters.
On the bad side Nextels are far from small but do take a beating and most people with a fashion sence consider them ugly. Persoanly I like my i95 it just works unlike the phones from any other provider (mostly Sprint and SBC)
USB 2.0 and enough bandwith for RAID? Lets see 480 Megabits a second divide by 8 for a whopping 60 Megabytes a second half duplex. I have single SCSI drives that do more than that. OK maybe it's reliability your after I dont think anybody has ever written multipath for USB though it could work with existing Mulipath drivers so we will give it the benifit of the doubt. So you need something with a pile of USB root hubs on a PCI-X card.
Now if you dont care about speed and just realibility maybe a little hardware mirror on a USB to IDE bridge chipset and pray the chipset doesent fail.
I'm sorry I'll keep my Fiberchannel and firewire USB is for cheap toys thats about as far as it can go without the ability to multiple initiators (someting every high speed external bus but USB has least that I can think of on a modern PC) flame on you USB zellots:) (I mean that in the nicest way)
Ah the real problem peo;le keep on saying welll we were at nearly peak demand and then somethng broke causing a cascade failure well gee some of us are engineers and how do we stop deband based cascade failures overengineer the problem.
On a small scale I see this all the time in datacenters circuts running at 80% or above asking for a breaker to trip and this is at seady load. Throw a hard problem at them like running distributed.net throw a pile of dd if=/dev/sda of=/dev/null and generaly tax the machines out and often you will see that circut run at 120% and trip out, bang a bunch fo low end servers offline. Reset the breaker and if they are set to autopower on watch it happen again sometimes. The first rule of engineering for reliability s never tax out any segment. If your cheap stay N+1 redundant it's not that expensive if your serivces are critical stay N*2 redundant. Our power needs are becomming more and more critical our power grig needs to become more reliable this means more power generation and transmission at the bank end.
That coupled with the fact that while inconvient the eletrical systems alone does not make a grea t target for a terroist. Drop planes out of the sky kill people etc is an exersize in terror blackups are an exersize in annoyance. Anything thats important has backups in place.
Yea we have power outages they are generaly the fault of the eco freaks fighting tooth and nail about doing any sort of improvment to the power grid democracy is good for one thing nothing gets done fast unless you put it on tv and spin spin spin.
Actualy those skiller engineers come from a price to performance thats probably different than some of us would like. They fixed all the obvious single point s of failure thats a good first step. Pretty much it's a question of disversifying things so there isn't one big tower to knock down and loose major percentaged of transmision ability. They didn't look at the relitive centralization of power generaton especialy with the Nukes in CT being shut down. We have at best a N+1 system those are ok for general use but moving to something thats closer to a proper (n+1) + (n+1) systems thats discreat untill that last possible moment it's plausable unless we take the generation and bring it closer to the use point.
Yes you have th fundemental problem but you dont see it why is software different from a tangable good? Or maybe I'm missing the point that companies would rather lease you everything and not let you buy anything thats not a consumeable. Is that realy the furture to look forward to where you cant own anything you just rent? Sorry I must put MegaCorp Gas in my MegaCorp car or I void my lease agreement and would have to pay a fine as Megacorp's gas has a special tagging agent that the car detects and allows it to run?
I guess you want a Cisco IDS tied to a Pix with shunning turned on? SNORT does one thing well detect nasty packets and flows it's then up to you to do something about it in an automated manner. A little scripting can generaly get this done.
No server is immune to Hacking it's just a question of ease. Microsoft servers are often not well configured and often not maintained. They are also a large percentage of the host avalible on the internet so a good target. This all cuppled with a apparent pollicy that ease of use is more important than security by default from Microsoft has led to the current situation. The more secure you make a box the harder it is to work with. Linux, BSD and Unix in general are somewhere in the middle by default. You could move to secure versions that are a pain to do anything on and I do mean a pain. Just think of it as a variable rather than a booleen it's not is it secure yes no but rather how secure is it and allways remember if effort is applied by a knoledgeable expert it can become more secure to a point.
Slashdot Mirror
Acutlay if you ever looked at how DSL works smalls towns take a one time cost to increase bandwith. DSL works as follows:
Home "Modem" takes Ethernet frames stips off the ethernet bits and adds ATM headers at layer 2 and throws it onto the actual DSL line.
The DSLAM general owned by the incumbant telco taks the DSL signaling and puts it into the ATM cloud.
The allready exisitng ATM out of town infrastructure they backhaul this to whever the DSL providers decide to hook up and via whatever sized pipe the provider wants. Because it's ATM if it's a garenteed speed they have to have the pipe to fit it in. The ATM cloud is the only parts the telco has to deal with expanding it's realy one time costs of getting faster line cards as they own the backhauls anyway to support more bandwith. Technicaly because of the way ATM works there cant be bandwith issues in the bachhaul portion because ATM uses two basic types of PVC there is CBR thats fixed rate and EVERYTHING reserves the space for it or VBR that is as allowed. Most DSL I have seen gets provisioned at VBR with a fixed max and min speed all the min speeds added up have to be smaller than the all the connection from point a to b have room for or the PVC wont come up and will never work at all.
Small telcos general have the issue of no close points to backhaul to cheap. If I get a line into New York or most major cities I can get bandwith for $10 a megabit and up with good bandwith at under a hundred. But you have to get the data there first to a rural ISP that might be a long backhaul from some small city thats outragious at the small end.
OK I will have to disagree with the fee structure so far at least in Connecticut I work with several small and midsized ISP's (regional) the DSL fee structure as it stands is in the telcos favor. First off there is a fairly steep ATM access price thats equivilent to a leased line and then some generaly in the neighborhood or 30-45k for an OC3 ATM monthly with a good 100k install fee. Then they want 35 per line plus an install fee. General hassels getting service and support that you see with there own service but not to the same extent. Now on top of this they are offing a 29 a month contract for a year agreement with a waived install. Now if this wasent bad enough they have 11 a month in nebulus DSL access fees charged directly to the users bill. All tole they are making 46 bucks a month per DSL line from a competitor while selling theres for 40. This is clearly a monopoly abusing it's power by subsidising it's DSL and internet server that runs at a loss via revenues from it's traditional services. Funny thing with all of this is if I orded a dry pair and didn't have to use there DSLAM's I could provision the line for 7 a month and the costs of supporting a DSLAM in the switch house or someplace very near by it possibly on the POLE outside (Space rented from the Eletric company pretty inexpensive actualy). I could also backhaul via my own circuts and skip the nebulus ATM cloud and those associated provisioning costs (Doesnt anybody relize ATM is dying a death it deserves it's a dead end technicaly, it's easier to throw more bandwith at the problem than ration everything out cheaper as well)
Having the local government run a copper last mile plant might be a good idea. If you ever have used low density copper it's realy not that hard to maintain unless somehting falls on it you might not deal with it for years. A hybrid fiber and copper plant might be a better idea it's not that hard to convert and or split out to copper. Copper is still just a lot more useable that fiber into the home and getting a few thousand feet of fiber put in it's a lot cheaper and easier that miles worth if you realy want fiber. A side note on fiber it's possible to do it on the pole but cant we finialy get away from poles and stay burring things where they belong?
Whats wrong with a router doing NTP all the routers I configure use it and most of them serve it up as well. This was the same argument people made about not puting DHCP on the router again something thats very usefull to have in place. Yes it's a trade off vs centralized management but in the case of an NTP server what do you manage centraly?
It was bad form for Netgear to use somebody elses NTP server. It would have been more apropriate to have it use the DHCP flag (for those cable modems) for NTP server and use that one if avalible otherwise use some DNS name from netgear.
Ah but she could make an argument that she might be opening herself up to criminal charges by having the information disclosed. You cant use a civil proceding to get around due proccess it what could be a criminal matter or due process is worthless.
I beleive you want an Nforce2 chipset it's not socketed but todays new PC owner dosent change out there video card. But otherwise slots are usefull for neatness sake if nothing else and high speed busses. The fastest external connector is firewire 800 at 100MB a sec half duplex (Not entirly sure ont hat bit) It cant deal with GigE speeds. PCI at base does 132MB a sec and can clock much faster and wider currently to around 1024MB a sec thats 4 GigE adapters running full out to get buss saturation. And most systems that implement PCI-X run multiple buses of it for more potential speed. Firewire 800 is porbably fast enough for nearly everything a home users needs to do except gigabit networking but dosent scale to servers.
Define work? With current routing topologies you take down all the tier ones and your not getting out of the USA and will have trouble getting much farther than that. Contract wise the tier ones have been applying a lot of presure on the tier 2 guys not to advertise interconnects and often have good reasons not to. Add to this the massive ammounts of long haul centralization take out a few MAE points and things would be bad VPN's are replaceing the long haul circut and as it gets nastier and naster out there firewalls are the norm blocking trafic through corps private backbones and satalite links.
So does this mean they can send adds for the cheapest booze in town? Or for eddies Cardboard box emporium? If your going to spam somebody it's generaly commercial and thats the one group thats not interested in the indigent.
As a hiring manager PHD's arent particulary atractive in and of themselves. With the poor state of pratical education persoanly I would rather hire a person with more experience than an over educated one. The only roles that generaly seem to be looking for PHD's are CTO and senior research positions and either of those again want to see a lot of expereince.
More importantly in NJ the casino's can still ask you to leave if they think your counting cards as they do not require a reason to do so.
OK I dont know where you are or even if your in the USA. But lets not let that stop me making the usual american twist on things. First off in a lot of states even eletrical engeineers arent allowed to work on commercial or industrial wiring it has to be a licened eletritian otherwise a nice building inspector can come in and close the building period end of sentance yes the nice sherif will escort you out now till it's fixed and up to code. Now with that all said rewiring a mess like you describe is anywhere from easy to very complex depending on your UPS and gen set needs. From your description of the main fuses (those nice 1liter bottle looking things) you are running a decent ammount of current the perfered method would be to get a new pannel mount it get your eletric company out to move over the mains while you move over all the circuts. If at all possible oversize the pannel and ask for an increase in incoming amps they never hurt. If you realy need minimal downtime try and get a new main service attached it to the new pannel and move over circut but circut to the live new pannel it's not that dangerious just stay slow and pedantic.
So your saying your compnay had the usual poor design of an M&M hard on the outside soft on the inside. Realy at least they should require corperate standards for laptop firewalling and virus scanning at minimum.
The point is to allow other documenters to make the more human readable source. The definitive technical documentation either has to come first or the origional expert aka the program team needs to make the human readable for dummy's type then write the hard core technical documentation and thats backwards. Your asking probably the wrong person or group to write the dummed down documantation when what should happen is to have an outside interpeter write the dummed down version after working through the expert material.
I mean that first tier documentation should be exceedingly technical this is what the programmer the programming team or possibly the tech writter assigned to such groupings should be writting it. Programmers dont write good manuals in general but they do need to cover the specifics in detail these things are often very cryptic to everybody that hasent been around the particular piece of technology before. But this is the level of documentation required for others to be able to realy use the application.
My general issue with for dummy's books is lack of technical detail they are not clear enough to limitations. Take for example two simple lines talking about remote systems connected to a server application here is the dummy's version at best:
The server can support hundred of users on a modest server.
A better version:
The server utilizes 1 48KB desciptor bound to OS memory block size per active session. The session unique identifer is arbitrarily a unsigned int as defined on the local system.
One just gives you no real information the other gives detained sizing requirements as to the session overhead and max count. Generaly you dont read technical manuals just as you dont ready encyclopedia's (not a great example) For dummy's books are arguably educational program documention should be referance.
I dont hate users I dont like users I generaly dont deal with users thats what help desk is for. If by users you mean System Admins I deal with them often mostly by fixing there issues.
Oh dont worry I make plenty of money working with windows as once people figure out the paper MSCE's cant do what is required they hire a compotent consultant. :)
BTW I like the right tool for the problem beleive it or not clusters of Linux boxes dont work for any problem yet.
Yes documentation should be the expert on the topic written for somebody with a background in real engineering (your average MSCE dosent count) let the howto's and the for dummys books deal with spoon feeding cookbooks to end users if your having the authoritive person on the subject write documentation aka the programming team write the most technical documentation you should ever need without having to do redo code yourself.
I say this because there are to many porly documented applications out there. Documentation to often is looked at by the marketing department and dumbed down so nobody might get scared of it. If you have ever looked at the home service manual for a Saturn (the $500 one thats an option) that nearly would allow you to machine replacment parts thats documentation. Want something easy to read with pretty pictures get a for dummy's book aka the dumbed down book from somebody that read and understood most of the documentation.
Why bother with wastefull wind generation plants (they waste space and generate noise) Fuel cells work can be powered easily off biomass (aka our waste) and with a unit in every house producing 5kw we would have a good deal of excess capacity while keeping generation local the unit sits on a pad outside and is the size of a refrigerator. Realy is you use a pump to compress the natural gas to a storage tank you have a decent backup supply.
For the more power consious attach a battery farm in the middle (ok the eco people dont like these to much but they are reliable) with a larger inverter it could handle larger short peak loads. You still have a power grid to share power from house to house again filling individual peaks (eletric ovens on clean eletric water heaters) Obvious those 50s all eletric houses would need bigger units (or to get a life and forget about eletric heating with the current trends away from nucular power)
OK having worked at a company that everybody had nextel phones they are a great time saver with PTT. You just chirp in as a manager and ask your questions often from meetings if something came along that you werent sure about without having todrag the person there. The big thing is most home users arent responcible enough to turn off the PTT speaker when it's not usefull like in the movies. Add to that that PTT minutes were unlimited on a lot of plans meant a lot of people spent to much time instead of making a phone call. Persoanly Nextel's service is a lot better than Verison simply because it's the only company that realy was pushing to corprate users. Go into a 3rd story underground datacenter from most ISP's that didn't have an affiliated cell phone division and Nextels would work do to repeaters.
On the bad side Nextels are far from small but do take a beating and most people with a fashion sence consider them ugly. Persoanly I like my i95 it just works unlike the phones from any other provider (mostly Sprint and SBC)
USB 2.0 and enough bandwith for RAID? Lets see 480 Megabits a second divide by 8 for a whopping 60 Megabytes a second half duplex. I have single SCSI drives that do more than that. OK maybe it's reliability your after I dont think anybody has ever written multipath for USB though it could work with existing Mulipath drivers so we will give it the benifit of the doubt. So you need something with a pile of USB root hubs on a PCI-X card.
:) (I mean that in the nicest way)
Now if you dont care about speed and just realibility maybe a little hardware mirror on a USB to IDE bridge chipset and pray the chipset doesent fail.
I'm sorry I'll keep my Fiberchannel and firewire USB is for cheap toys thats about as far as it can go without the ability to multiple initiators (someting every high speed external bus but USB has least that I can think of on a modern PC) flame on you USB zellots
Ah the real problem peo;le keep on saying welll we were at nearly peak demand and then somethng broke causing a cascade failure well gee some of us are engineers and how do we stop deband based cascade failures overengineer the problem.
On a small scale I see this all the time in datacenters circuts running at 80% or above asking for a breaker to trip and this is at seady load. Throw a hard problem at them like running distributed.net throw a pile of dd if=/dev/sda of=/dev/null and generaly tax the machines out and often you will see that circut run at 120% and trip out, bang a bunch fo low end servers offline. Reset the breaker and if they are set to autopower on watch it happen again sometimes. The first rule of engineering for reliability s never tax out any segment. If your cheap stay N+1 redundant it's not that expensive if your serivces are critical stay N*2 redundant. Our power needs are becomming more and more critical our power grig needs to become more reliable this means more power generation and transmission at the bank end.
That coupled with the fact that while inconvient the eletrical systems alone does not make a grea t target for a terroist. Drop planes out of the sky kill people etc is an exersize in terror blackups are an exersize in annoyance. Anything thats important has backups in place.
Yea we have power outages they are generaly the fault of the eco freaks fighting tooth and nail about doing any sort of improvment to the power grid democracy is good for one thing nothing gets done fast unless you put it on tv and spin spin spin.
Actualy those skiller engineers come from a price to performance thats probably different than some of us would like. They fixed all the obvious single point s of failure thats a good first step. Pretty much it's a question of disversifying things so there isn't one big tower to knock down and loose major percentaged of transmision ability. They didn't look at the relitive centralization of power generaton especialy with the Nukes in CT being shut down. We have at best a N+1 system those are ok for general use but moving to something thats closer to a proper (n+1) + (n+1) systems thats discreat untill that last possible moment it's plausable unless we take the generation and bring it closer to the use point.
Yes you have th fundemental problem but you dont see it why is software different from a tangable good? Or maybe I'm missing the point that companies would rather lease you everything and not let you buy anything thats not a consumeable. Is that realy the furture to look forward to where you cant own anything you just rent? Sorry I must put MegaCorp Gas in my MegaCorp car or I void my lease agreement and would have to pay a fine as Megacorp's gas has a special tagging agent that the car detects and allows it to run?
I guess you want a Cisco IDS tied to a Pix with shunning turned on? SNORT does one thing well detect nasty packets and flows it's then up to you to do something about it in an automated manner. A little scripting can generaly get this done.
No server is immune to Hacking it's just a question of ease. Microsoft servers are often not well configured and often not maintained. They are also a large percentage of the host avalible on the internet so a good target. This all cuppled with a apparent pollicy that ease of use is more important than security by default from Microsoft has led to the current situation. The more secure you make a box the harder it is to work with. Linux, BSD and Unix in general are somewhere in the middle by default. You could move to secure versions that are a pain to do anything on and I do mean a pain. Just think of it as a variable rather than a booleen it's not is it secure yes no but rather how secure is it and allways remember if effort is applied by a knoledgeable expert it can become more secure to a point.