I use dsl.net - they're more business grade but will still deal with consumers. They gave me as many static IP's as I could justify and seem to have no problems with how many things I attach to the line.
Unless it's a kernel issue there's no reason why they should have to reboot. One of the nice things about unix is that you can fix/replace/change parts of the system without affecting others. If they're going to update their popd why is that grounds for a reboot? Move the old binary, stick in the new and restart the daemon.
Was the story about you embarrasing a Microsoftie at a conference true? Specifically, that he was insisting that their implmentation of ksh in their unix compatibility kit was true to the "real" thing and trying to argue the point with you. The argument ended when somoene else finally stood up and informed the speaker who he was arguing with.
This is normally called "EtherChannel" in cisco-speak. It's usually implemented at the adaptor level and requires that you be connected into a switch that supports it. I'm not sure about support under Linux or other free unixes but the card manufactures support it under windows. Intel has support for it- I believe they call it "Adaptor Teaming".
One caveat about bonding multiple connections -- unless you're doing something like MultilinkPPP that "sprays" packets out you're not going to see a bandwidth increase on a single connection just by adding multiple pipes. In other words, if you have 2 T1's to the internet you're not going to see 2x1.5Mb/s on a single connection. By default a cisco router will load balance on a srcdst ip basis over each link.
Gene matching is very data intensive -- basically one of the things that they're going to be doing is matching each part against every other part to see where things match. This means having to have a copy of all of the data available to each client. Seti/distributed/etc are really just compute-intensive apps -- they don't need to see a huge data set to do their work.
Not the same at all - the MPAA causes the sited to become unavailable to anyone. The RBL is an opt-in service and if you don't like the way they run things you're free to ignore MAPS and access the spammer sites anyway. Your ISP uses MAPS? Fine, find a new ISP or make a case to your current one..
Schools are getting picky. Most won't accept non-working hardware, and many won't accept less than a devent Pentium-class machine. 486's aren't going to fly -- they have to dispose of them too, rememeber?
My wife take classes during the day and works in the evenings. My flexible hours allow me to spend mornings with her and work during roughly the same hours. I can come in at 10:30AM and leave at 8.
Disadvantages are that dealing with others on more normal schedules can be a pain.
Why not ship good old DOS with the box? It's gotta be cheap -- it's moderately usefull (you can at least use it to rawrite boot floppies), and would not subject the manufactures/vendor to any undue support issue like shipping a linux distro would.
Haven't we been arging all aong that ISP's should be considered Commmon Carries -- becauase being so makes them less liable for content -- and hence less likely to have to respond to a court order to block certain things.
What's the *right* thing here? I don't think we can have our cake and eat it too.
Another angle -- I don't think that this really says that all ISP's are not common carriers. I'm reading more that AOL isn't a common carrier becuase it provides unique services and content. a "real" ISP is only providing the bandwidth and certain infrastructure - i.e. DNS (akin to 411)
The problem with this is that there are non-porno sites that are being blocked and a person would have to ask for. Would someone who is researching AIDS because they or a family memeber was just diagnosed with it really want to tell a librarian this? There's no such thing as librarian-customer privelege to prevent disclosure of such information. In a small town this could be a major issue -- everyone knows everyone and news travels.
Not worth the effort. One traceroute to the server and finding the physical server is irrelevant -- all you need on a good network is to find the closest router to the box, use the ARP tables to find the MAC address of the box, and then look at the switch to see what port it's connected to.
Actually I think that being able to pay a monthly fee to access any and all of the Beeb's audio archives would be excellent. Not too unlike a video store membership.
Re:Failure to implement open standards.
on
Linux Failover?
·
· Score: 1
Actually, my F5 Labs BigIP load balancers suppport VRRP also. Works rather wall too I might add. I'm actually running Redundant Border routers (cisco 3620's) with HSRP (cisco's version of VRRP), feeding two Nokia IP440's with VRRP on all interfaces, and behind them are a pair of BigIP's also using VRRP on both the interior and exterior nets.
The data processing problems with the Sequencing part of the process are mostly data storage ones -- these sequencers just generate a *lot* of data and you need to keep track of it all.
The real crunching comes during the Assembly process. This is where all of the fragments that were sequenced prett randomly (they call this Shotgun Sequencing for a good reason) are fed thru algorithms that take and try to reassemble all of the parts into the finished product like big jigsaw puzzle where the pieces don't really key in, but sort of overlap. So far the algoritms to do this realy on having an extrememly large contiguous memory space to be able to do this. I do know that they're hard at work on rewriting the algorithms to do this in a distributed manner but when you've got money to burn on throwing hardware at the problem what's the point? In any case, the HGP is already a sort of distributed processing setup -- labs all over the country are doing independent sequencing and then combining their data. Once their sequencing is done they'll then face the same problems as anyone else of needing to Assemble all of the fragments -- and this would be the perfect place for a distributed algorithm.
For the old time COBOL monkeys out there, primeOS.com is available. I think there are ~6 prime boxes still in operation, so there's a built in audience.
5. I just shoved mine out the door last month. Actually it's probably still in operation -- I gave it to the guy who did our administration for it. CVSI (www.cvsi.com) who bought the remains of Prime actually owns prime.com.
Apparently the government still uses a bunch of them. The FBI specifically I believe... Imagine that...
Why would a recruiter *not* be calling? They only get paid when they place someone - the more people that the can shove thru the pipeline the more chances they have of someone getting hired and causing them to get a fat recruiting fee.
Slashdot Mirror
I use dsl.net - they're more business grade but will still deal with consumers. They gave me as many static IP's as I could justify and seem to have no problems with how many things I attach to the line.
Unless it's a kernel issue there's no reason why they should have to reboot. One of the nice things about unix is that you can fix/replace/change parts of the system without affecting others. If they're going to update their popd why is that grounds for a reboot? Move the old binary, stick in the new and restart the daemon.
Just curious...
One caveat about bonding multiple connections -- unless you're doing something like MultilinkPPP that "sprays" packets out you're not going to see a bandwidth increase on a single connection just by adding multiple pipes. In other words, if you have 2 T1's to the internet you're not going to see 2x1.5Mb/s on a single connection. By default a cisco router will load balance on a srcdst ip basis over each link.
Gene matching is very data intensive -- basically one of the things that they're going to be doing is matching each part against every other part to see where things match. This means having to have a copy of all of the data available to each client. Seti/distributed/etc are really just compute-intensive apps -- they don't need to see a huge data set to do their work.
Not the same at all - the MPAA causes the sited to become unavailable to anyone. The RBL is an opt-in service and if you don't like the way they run things you're free to ignore MAPS and access the spammer sites anyway. Your ISP uses MAPS? Fine, find a new ISP or make a case to your current one..
Uh, *what* linked page?
Why bother downloading either of them? Do an FTP install and just take what you need. You're sure to get the latest this way.
Schools are getting picky. Most won't accept non-working hardware, and many won't accept less than a devent Pentium-class machine. 486's aren't going to fly -- they have to dispose of them too, rememeber?
My wife take classes during the day and works in the evenings. My flexible hours allow me to spend mornings with her and work during roughly the same hours. I can come in at 10:30AM and leave at 8.
Disadvantages are that dealing with others on more normal schedules can be a pain.
Why not ship good old DOS with the box? It's gotta be cheap -- it's moderately usefull (you can at least use it to rawrite boot floppies), and would not subject the manufactures/vendor to any undue support issue like shipping a linux distro would.
What's the *right* thing here? I don't think we can have our cake and eat it too.
Another angle -- I don't think that this really says that all ISP's are not common carriers. I'm reading more that AOL isn't a common carrier becuase it provides unique services and content. a "real" ISP is only providing the bandwidth and certain infrastructure - i.e. DNS (akin to 411)
I got mine via my Wired subscription. Even the box is co-branded with Wired...
The problem with this is that there are non-porno sites that are being blocked and a person would have to ask for. Would someone who is researching AIDS because they or a family memeber was just diagnosed with it really want to tell a librarian this? There's no such thing as librarian-customer privelege to prevent disclosure of such information. In a small town this could be a major issue -- everyone knows everyone and news travels.
Not worth the effort. One traceroute to the server and finding the physical server is irrelevant -- all you need on a good network is to find the closest router to the box, use the ARP tables to find the MAC address of the box, and then look at the switch to see what port it's connected to.
Didn't Apple vs. Micro$oft tell us how this movie ends? Look and Feel isn't protected.
Actually I think that being able to pay a monthly fee to access any and all of the Beeb's audio archives would be excellent. Not too unlike a video store membership.
Actually, my F5 Labs BigIP load balancers suppport VRRP also. Works rather wall too I might add. I'm actually running Redundant Border routers (cisco 3620's) with HSRP (cisco's version of VRRP), feeding two Nokia IP440's with VRRP on all interfaces, and behind them are a pair of BigIP's also using VRRP on both the interior and exterior nets.
The real crunching comes during the Assembly process. This is where all of the fragments that were sequenced prett randomly (they call this Shotgun Sequencing for a good reason) are fed thru algorithms that take and try to reassemble all of the parts into the finished product like big jigsaw puzzle where the pieces don't really key in, but sort of overlap. So far the algoritms to do this realy on having an extrememly large contiguous memory space to be able to do this. I do know that they're hard at work on rewriting the algorithms to do this in a distributed manner but when you've got money to burn on throwing hardware at the problem what's the point? In any case, the HGP is already a sort of distributed processing setup -- labs all over the country are doing independent sequencing and then combining their data. Once their sequencing is done they'll then face the same problems as anyone else of needing to Assemble all of the fragments -- and this would be the perfect place for a distributed algorithm.
5. I just shoved mine out the door last month. Actually it's probably still in operation -- I gave it to the guy who did our administration for it. CVSI (www.cvsi.com) who bought the remains of Prime actually owns prime.com. Apparently the government still uses a bunch of them. The FBI specifically I believe... Imagine that...
Why would a recruiter *not* be calling? They only get paid when they place someone - the more people that the can shove thru the pipeline the more chances they have of someone getting hired and causing them to get a fat recruiting fee.
The short story on the nickname is here.
Actually it hit a high of 90 so far...
Look here
For the Unix impaired I'm assuming that you are referring to "talk" and "finger".
Legato actually has a Networker module out for teh Filers now.