Recording live sets with digital media can be quite a pain in the... challenging. Make sure you have 1-2 analogue compressor/limiter per channel before the signal hits your deck. Digital distortion is nearly non-restorable and will ruin the recording beyond usability.
Because of this I routinely record on analogue 4track (Tascam 424Mk3) and feed that into my PC for postprocessing. In fact, most of the time I do the mixdown on the analogue 4track and only record the pre-mixed stereo into the PC. Saves time and hassle.
So where do you want the 4track for? If only for home studio use, a decent software (CoolEdit, Cubase) will be much more convenient and more flexible to use.
The Symantec firewall formerly was known as "Raptor Firewall" or "Axent Raptor Firewall". It is a hybrid firerwall with quite a number of transparent security proxies, whereas Linux machines "only" do stateful plus maybe (standard) proxies for only a limited number of protocols. For a class overview see http://wyae.de/secure_gateway/gateways.php
In my experience the Raptor is(was) quite good and not really comparable to a custom linux machine or off-the-shelf linux firewall (e.g. Astaro) - though I like the latter, too. It's playing in a completely different (IMHO higher) class.
The Raptor's SPs are among most stringent I know of - but can be a real pain to pass through for nearly-compatible stuff. The Notes SMTP gate was infamous for being rejected by Raptor because of RFC-noncompliance...
Apropos "maintenance-free": no forewall is maintenance-free. Never. You'll always have to have a look at the logs, at unusual behavious, etc. The only difference here is wether you have to care about building software patches yourself or to have a company do that for you. But the load of necessary maintenance work still is to be done. If you ignore that, you'll pay the price, probably earlier than later...
If you have less than 12 MB RAM (but more than 5), it is easier to start with Debian 2.1 (Slink) as kick-off base and then dist-upgrade to Potato (2.2) or Woody (3.0). You won't have to trick around the memory limitation. Using a PC with less than 8 MB does not really make fun, though.
If you only have a small harddisc, you should not use TASKSEL, but install a very basic system. AFTER that select the packages you really need. You might run into problems even with the base syetem when using very small (notebook) HDs, i.e. smaller than 200MB.
I am using Debian on old 386/486/K6's as well as new P3/P4's, so I might be biased... ;-)
Indeed - and it depends on what your cluster is supposed to be beneficial for. Ideal for "number crunch" clustering are tasks that require low bandwidth and high CPU performance - like movie rendering or testing alternative simulation parameters. For the latter projects like SETI@home, distributed.net or Folding@home have become famous. Most CPU work, neglectible network load. For SETI@home I have an average network throughput of ~50 bit/second. To saturate a 100Mbit/s network (not even switched) with SETI@home you'll need approx. one million (1.000.000) PCs.
As for network - do you need throughput or low latency? Depeding on your problem small changes in algorithm can do wonders. E.g. for film rendering you might choose a few NAS and a hoard of dumb/diskless rendering slaves. If you copy the model libraries (for the included figures, textures, etc.) onto a local disk at the beginning of a scene render run, you will decrease net load a big deal (I've done that with Provray rendering myself).
If you don't have the rerssources to buy e.g. Myrinet, try alternative architectures if they might fit your problem, e.g. hypercubes (see other posts) or models like Flat Neighbourhood.
Basically the security is the same as of a directed RF link. Even worse, as RF does not scatter well.
Put *something* into the beam - anything, like a window (best with some dirt/dust), dust or mist in the air. The light scattered off (a few percent) will still be enough to be detected with standard telescopes even from a distant observer (distant = easier to hide).
So: just a marketing hype. The old common-sense still has to be applied: if you don't absolutely control the transfer media (e.g. in-house cable), you are susceptible to eavesdropping.
Wrong idea: Translation still is protected
on
Abusing the GPL?
·
· Score: 1
By replacing names etc. with nonsense words you only create a translation. Translation of copyrighted and -lefted works is subject to permission (i.e. license) of the copyright/-left owner.
Another view: the important part of computer programs is the structure, not the actual code (text) representation. You simply copy the protected essence - and thus violating any license or copyright law.
So if using a gooblegrocked code fragment you are violating licenses twice: once when "translating", second when using the translated code.
S/MIME seems to be no longer supported in current Netscape (version 6.0 and newer). There were problems transfering signatures/mails between mail programs of different brands (i.e. M$ and Netscape) with S/MIME. PGP / GnuPG is without any problems for any mailprogram when using the tray application.
Fault-detecting switches (Alteon, Big5, Cisco Local Director) do transparent switching - as do software HA solutions (Rainfinity, Linux Virtual Server).
BUT...
...when the connection is being switched to a new server (for load balancing or failure switchover), the server's application does not know the connection and thus the switched one will be rejected. You will need to program your application to keep and distribute state of the connection. If you do stateless applications (e.g. web-based), then at least the (next) IP-connection will have a clean switchover on the srever side.
This is why for backups you only should take tarballS - just TAR, no gzip, bzip or compress option. Tar is designed to jump bad blocks, this is why you still have the blockish design of the data stream (see manpage). So if you do compression all data behind that spot will be corrupt.
Short: *.tar for backup, *.tar.gz (or similar.Z or.bz2) for transfer.
...if I my interpretation of the description is correct. They basically seem to break down each image into smaller bits and assemble them later - and only transmit the differences. So the textualized representation may read something like
first frame
left up (pos 0,0) is a 16x32 block, near black (rgb #111111).
next to it (pos 16,0) a 16x16 block, grayish (rgb #112211)
below that a block (pos 16,16) with 16x8 green-grayish (rgb #115511)
below that a block (pos 16,24) with 16x8 block, greenish (rgb #05BB05)
next frame (Logo background appears in the middle)
block change in middle (pos. 8,8), size 16x16, black (rgb #000000).
next frame (Logo starting with bright expanding spot)
block change in middle (pos. 16,16), size 1x1, white (rgb #ffffff).
next frame (dito)
block change in middle (pos. 14,14), size 4x4, white (rgb #ffffff).
...etc...
Something like a poor man's MJPG+MPEG. Maybe, if not using fix colors but linear gradients (4 values total = left-right and up-down) the quality can be a bit better.
OTOH this compression is designed for mini-screens with waaaay sub-optimum quality anyway, so blockish compression is not an issue here?
A close look at a demo and the algorithms would be interesting, agreed.
T-Rex (according to the docs I gobbled) is a set of specifically configured proxies, notably Squid. This is definitely not comparable to a packet filter as PF, IPF (both *BSD) or IPchains, IPtable (both Linux).
AMD, Cyrix and Transmeta do x86 compatibility and play catchup with Intel's x86 development.
Sun (Sparc), HP (HP) and IBM (PowerPC) do their own server & workstation chips as usual - on a very low level (counting CPUs here), as usual.
According to my CPU statistics Intel is the singular heavyweight with respect to CPU development here?! So Intel aquiring one of his (very) few competitors is a Bad Thing(tm) with respect to a healthy architecture diversity.
I got a used (=cheap) DAT/DSS tape that works reliably for me. It only takes 2 GB (uncompressed) per cartridge, but a (90m) cartridge only comes around $3. Yes, that is about the same price range as CD-R, but tape is much more forgiving to load issues plus you do not need to change as often if you use e.g. GNU tar with the --multi-volume option. Plus the DATs are rewriteable/reusable.
As sidenote the top record on my old 233MHz PC (SCSI and IDE): backup plus harddisc-recording plus MP3-coding plus burning a CD - all without problems, well, except slow interface response...
Why DAT/DSS-1 (maybe -2) and not DLR, QIC or DAT/DSS-3? Simple: tapes and drives are much cheaper (single and per volume).
As someone else noted: make sure to TEST your backups. I myself use GNU tar with the option --verify for this successfully (and --multi-volume, as said above). Additionally I check occasionally with friends wether my DATs still are compatible (i.e. properly adjusted) with their DATs - just in case...
AFAIK Microsoft ASPs run - as (part of) the IIS - in kernel mode. If you have a bug causing an online loop the only way to stop that is to stop and restart the whole IIS (or sometimes the whole WinNT system).
This will slow the development down quite a bit. Plus the ASP/IIS/COM+/ODBCsomething can surprisingly easy develop a nondeterministic behaviour which is hard to understand and even worse to reproduce (read: restore after severe server crash).
In contrast to this standalone CGI (programs, not necessarily scripts) does not heavily interweave with other components but uses (or: should use) clean interfaces that are easy to understand and maintain. In addition to this standaloneable CGIs can be tested without any web service. First do an ASCII version of what you want to achieve, test it locally (like QUERYTYPE=POST; echo "parameter=locally&defined=from&command=line" |./yourcgibin)
and after that you can wrap it into nice HTML.
The main advantage of non-sinewave voltage conversion lies in the simpler setup of the conversion electronics. For rectangular you only need a switch and an inversion switch - then cycle off, on, off, reversed on,...
The major disadvantages of non-sinewave compared to pure sinewaves are:
worse efficiency - transformators work best with pure sine. Non-sinewave cause a number of useless spikes and inductive resistance.
unwanted & dangerous spikes - any non-smooth waveforms causes transformators or motors to create high-voltage spikes that can harm or even destroy your equipment
humming - with a pure sinewave you only have the standard AC frequency of 50/60 Hz. Other waveforms - especially rectangular - consist of a number of overtones (have a look at the waveform's FFT) that can make it easily through the filtzers (designed for 50/60Hz). This will be especially distracting for music equipment.
These BEAM robots - among them the featured one - are in most cases nonprogrammable, hardwired machines like photovores. While small, simple robots have a certain fascination, quite a number of them are extremely "dumb" (because hardwired). It can be quite a lot of fun to work with these, though.
If this is a must, brutal (and safe) methods like using high-capacity de-magnetizers, sanding off the platter cannot be used.
I guess you want to re-use the hard drive after deleting? Then you are stuck with software methods. A stated before you will need MANY truly random overwrites before the disk really is not recoverable. With Linux try (for drive/dev/hdX) something like below and repeat it 50 times. That should do the job.
Even if they had a tamperproof stream from first medium to speaker: nothing is easier than setting even the stereo boxes into a closed cabinet along with a decent microphone.
(Un)Fortunately our brain still has (copyable) analogue input and especially no bit scrambling - so all and each system share this kind of weakness.
You maybe remember the videotaped movie theater movies (StarWars E1, Martix, whatever)? Even with (oldfashioned) media control (movie reels) copy protection and pay-per-view (in the theaters) it was not a problem to get fairly decent quality.
Sure? I once witnessed an Win update (don't remember which) doing the update by removing the old files (known to be old Win versions) and copying in the new files. The (formerly) used (and removed) Microsoft driver was NOT among the ones listed with the new Win. Oops.
Yes, drivers are tied to hardware - but they are also part of a software system. And if that system is too good in removing its own "garbage", there is a problem.
I guess for OEMs business will be as usual. Just that they will be selling a "1 year subscribtion of newest BuzzWordWare(tm) software included" instead of "Newest BuzzWordWare(tm) software included". The only problem I see is that the hotlines will have a boost in calls during 12-18 months after the first rollouts, when the customers start complaining.
Then it depends how the "BuzzWordWare(tm)" company will be handling OEM rollouts - and rollout of the updates. Will the OEM hotlines be able to offer immediate help ("just keep on paying and we prolong your subscription for you") or will they have to forward that to the other company?
Another problem may become orphaned device drivers. What if the "BuzzWordWare(tm)" company decides that this or that driver is outdated (as has happened to some of the Win9x drivers)? Then all of a sudden the whole system will stop working - no chance of keeping on using the old driver. You have subscribed, so you have to update or stop using, remember?
...the date when Mir comes flaming down. Maybe they will give her a tiny boost in between to gain a bit of time for fruitless discussions. So I guess it will take a bit - let's assume april (with the boost).
2001-04-16 19:47:13
Rough guess on time - visibility estimated from parameters.
I (help desk) am on internet with address 9.9.9.9 (maybe hide-NATed, maybe not). Two users with problem are on PCs 10.10.10.10 and 10.11.11.11 (private, internal network). Both are hide-translated (N:1 NAT) to 2.2.2.2 on the internet side. So how can I connect to the first, how to the second one? Which IP address do I use for the one or other?
If you do 1:1 NAT (e.g. 2.10.10.10 and 2.11.11.11) you won't gain any security from NAT as you can still address each client individually - and thus leaving them vulnerable to DoS (flooding) attacks.
The "ASP" biztalk aside (that's just a funky workd used here for a webserver with request forms), these seem to be standard helpdesk companies which use a remote management tool (like VNC, ReachOut, pcAnywhere, CCM, etc).
You basically allow the help desk to remotely control your employee's client PCs. This means that - if you do not throw a dedicated line or big VPN solution - every single PC has to be addressable to internet. Even with a well designed firewall in place this introduces quite a risk for DoS attacks against the clients.
So if you open your company to be remotely managed by external help desks keep an eye on your security needs. Especially take care that you do not open a backdoor through the help desk system (your_net --> VPN --> help desk --> VPN --> other_net) and that liabilities and NDAs are safe and watertight for you.
Remotely managing clients can ease administration quite a bit (as posted before), but can become quite a hassle if the system does not work properly.
And make sure security does not become second. The admin tool either must require the user to confirm session shareing and enable him to immediately regain full control (by pressing a(ny) key) in case of problems - or enforce login as different user (which will log out the current user if he does not abort that with confirming a "really logout" popup box). Otherwise you will not be able to tell who did this or that possibly unpleasant change.
The concept is absolutely not new. Every FDDI ring topology could be used for/with this. Whereas Token ring only allows one single packet to circulate within the ring, this "new" concept allows, no, requires the ring to be filled up.
The "new" idea was not the ring as network topology nor the storage itself (has been done acoustically as stated above), but to coordinate client-server clusters with this. But clusters organized in ring-topology are not new either.
In the proposed topology the master server just injects "to-be-done" packets into the ring. The clients pick up (and remove) one packet from the ring each time they want to start crunching the next work packet. The other packets will be circulating the ring until solved.
Main problem is that one will be either wasting a lot of transmission capacity for idle data circulation - or be running into capacity problems. There is a reason why most detail work when designing clusters goes into designing the optimum network architecture for the specific problem...
Slashdot Mirror
Recording live sets with digital media can be quite a pain in the... challenging. Make sure you have 1-2 analogue compressor/limiter per channel before the signal hits your deck. Digital distortion is nearly non-restorable and will ruin the recording beyond usability.
Because of this I routinely record on analogue 4track (Tascam 424Mk3) and feed that into my PC for postprocessing. In fact, most of the time I do the mixdown on the analogue 4track and only record the pre-mixed stereo into the PC. Saves time and hassle.
So where do you want the 4track for? If only for home studio use, a decent software (CoolEdit, Cubase) will be much more convenient and more flexible to use.
The Symantec firewall formerly was known as "Raptor Firewall" or "Axent Raptor Firewall". It is a hybrid firerwall with quite a number of transparent security proxies, whereas Linux machines "only" do stateful plus maybe (standard) proxies for only a limited number of protocols. For a class overview see http://wyae.de/secure_gateway/gateways.php
In my experience the Raptor is(was) quite good and not really comparable to a custom linux machine or off-the-shelf linux firewall (e.g. Astaro) - though I like the latter, too. It's playing in a completely different (IMHO higher) class.
The Raptor's SPs are among most stringent I know of - but can be a real pain to pass through for nearly-compatible stuff. The Notes SMTP gate was infamous for being rejected by Raptor because of RFC-noncompliance...
Apropos "maintenance-free": no forewall is maintenance-free. Never. You'll always have to have a look at the logs, at unusual behavious, etc. The only difference here is wether you have to care about building software patches yourself or to have a company do that for you. But the load of necessary maintenance work still is to be done. If you ignore that, you'll pay the price, probably earlier than later...
If you have less than 12 MB RAM (but more than 5), it is easier to start with Debian 2.1 (Slink) as kick-off base and then dist-upgrade to Potato (2.2) or Woody (3.0). You won't have to trick around the memory limitation. Using a PC with less than 8 MB does not really make fun, though.
;-)
If you only have a small harddisc, you should not use TASKSEL, but install a very basic system. AFTER that select the packages you really need. You might run into problems even with the base syetem when using very small (notebook) HDs, i.e. smaller than 200MB.
I am using Debian on old 386/486/K6's as well as new P3/P4's, so I might be biased...
SETI@home,
distributed.net or
Folding@home have become famous. Most CPU work, neglectible network load. For SETI@home I have an average network throughput of ~50 bit/second. To saturate a 100Mbit/s network (not even switched) with SETI@home you'll need approx. one million (1.000.000) PCs.
As for network - do you need throughput or low latency? Depeding on your problem small changes in algorithm can do wonders. E.g. for film rendering you might choose a few NAS and a hoard of dumb/diskless rendering slaves. If you copy the model libraries (for the included figures, textures, etc.) onto a local disk at the beginning of a scene render run, you will decrease net load a big deal (I've done that with Provray rendering myself).
If you don't have the rerssources to buy e.g. Myrinet, try alternative architectures if they might fit your problem, e.g. hypercubes (see other posts) or models like Flat Neighbourhood.
Basically the security is the same as of a directed RF link. Even worse, as RF does not scatter well.
Put *something* into the beam - anything, like a window (best with some dirt/dust), dust or mist in the air. The light scattered off (a few percent) will still be enough to be detected with standard telescopes even from a distant observer (distant = easier to hide).
So: just a marketing hype. The old common-sense still has to be applied: if you don't absolutely control the transfer media (e.g. in-house cable), you are susceptible to eavesdropping.
By replacing names etc. with nonsense words you only create a translation. Translation of copyrighted and -lefted works is subject to permission (i.e. license) of the copyright/-left owner.
Another view: the important part of computer programs is the structure, not the actual code (text) representation. You simply copy the protected essence - and thus violating any license or copyright law.
So if using a gooblegrocked code fragment you are violating licenses twice: once when "translating", second when using the translated code.
S/MIME seems to be no longer supported in current Netscape (version 6.0 and newer). There were problems transfering signatures/mails between mail programs of different brands (i.e. M$ and Netscape) with S/MIME. PGP / GnuPG is without any problems for any mailprogram when using the tray application.
One option may be lunchbox type computers (e.g. http://www.lunchboxcomputers.com/). While not as sneaky/slim as current notebooks they have enough room for mods and special equipment (as discussed in a current thread http://slashdot.org/article.pl?sid=02/01/07/194520 4&mode=nested).
Fault-detecting switches (Alteon, Big5, Cisco Local Director) do transparent switching - as do software HA solutions (Rainfinity, Linux Virtual Server).
BUT...
...when the connection is being switched to a new server (for load balancing or failure switchover), the server's application does not know the connection and thus the switched one will be rejected. You will need to program your application to keep and distribute state of the connection. If you do stateless applications (e.g. web-based), then at least the (next) IP-connection will have a clean switchover on the srever side.
This is why for backups you only should take tarballS - just TAR, no gzip, bzip or compress option. Tar is designed to jump bad blocks, this is why you still have the blockish design of the data stream (see manpage). So if you do compression all data behind that spot will be corrupt.
.Z or .bz2) for transfer.
Short: *.tar for backup, *.tar.gz (or similar
- first frame
- left up (pos 0,0) is a 16x32 block, near black (rgb #111111).
- next to it (pos 16,0) a 16x16 block, grayish (rgb #112211)
- below that a block (pos 16,16) with 16x8 green-grayish (rgb #115511)
- below that a block (pos 16,24) with 16x8 block, greenish (rgb #05BB05)
- next frame (Logo background appears in the middle)
- block change in middle (pos. 8,8), size 16x16, black (rgb #000000).
- next frame (Logo starting with bright expanding spot)
- block change in middle (pos. 16,16), size 1x1, white (rgb #ffffff).
- next frame (dito)
- block change in middle (pos. 14,14), size 4x4, white (rgb #ffffff).
...etc...
Something like a poor man's MJPG+MPEG. Maybe, if not using fix colors but linear gradients (4 values total = left-right and up-down) the quality can be a bit better.OTOH this compression is designed for mini-screens with waaaay sub-optimum quality anyway, so blockish compression is not an issue here? A close look at a demo and the algorithms would be interesting, agreed.
T-Rex (according to the docs I gobbled) is a set of specifically configured proxies, notably Squid. This is definitely not comparable to a packet filter as PF, IPF (both *BSD) or IPchains, IPtable (both Linux).
Intel does x86, Itanium and ARM - and soon Alpha.
AMD, Cyrix and Transmeta do x86 compatibility and play catchup with Intel's x86 development.
Sun (Sparc), HP (HP) and IBM (PowerPC) do their own server & workstation chips as usual - on a very low level (counting CPUs here), as usual.
According to my CPU statistics Intel is the singular heavyweight with respect to CPU development here?! So Intel aquiring one of his (very) few competitors is a Bad Thing(tm) with respect to a healthy architecture diversity.
I got a used (=cheap) DAT/DSS tape that works reliably for me. It only takes 2 GB (uncompressed) per cartridge, but a (90m) cartridge only comes around $3. Yes, that is about the same price range as CD-R, but tape is much more forgiving to load issues plus you do not need to change as often if you use e.g. GNU tar with the --multi-volume option. Plus the DATs are rewriteable/reusable.
As sidenote the top record on my old 233MHz PC (SCSI and IDE): backup plus harddisc-recording plus MP3-coding plus burning a CD - all without problems, well, except slow interface response...
Why DAT/DSS-1 (maybe -2) and not DLR, QIC or DAT/DSS-3? Simple: tapes and drives are much cheaper (single and per volume).
As someone else noted: make sure to TEST your backups. I myself use GNU tar with the option --verify for this successfully (and --multi-volume, as said above). Additionally I check occasionally with friends wether my DATs still are compatible (i.e. properly adjusted) with their DATs - just in case...
This will slow the development down quite a bit. Plus the ASP/IIS/COM+/ODBCsomething can surprisingly easy develop a nondeterministic behaviour which is hard to understand and even worse to reproduce (read: restore after severe server crash).
In contrast to this standalone CGI (programs, not necessarily scripts) does not heavily interweave with other components but uses (or: should use) clean interfaces that are easy to understand and maintain. In addition to this standaloneable CGIs can be tested without any web service. First do an ASCII version of what you want to achieve, test it locally (like QUERYTYPE=POST; echo "parameter=locally&defined=from&command=line" | ./yourcgibin)
and after that you can wrap it into nice HTML.
Did I mention that I like independence?
The major disadvantages of non-sinewave compared to pure sinewaves are:
- worse efficiency - transformators work best with pure sine. Non-sinewave cause a number of useless spikes and inductive resistance.
- unwanted & dangerous spikes - any non-smooth waveforms causes transformators or motors to create high-voltage spikes that can harm or even destroy your equipment
- humming - with a pure sinewave you only have the standard AC frequency of 50/60 Hz. Other waveforms - especially rectangular - consist of a number of overtones (have a look at the waveform's FFT) that can make it easily through the filtzers (designed for 50/60Hz). This will be especially distracting for music equipment.
Pure sinewave UPS come with a cost, though...These BEAM robots - among them the featured one - are in most cases nonprogrammable, hardwired machines like photovores. While small, simple robots have a certain fascination, quite a number of them are extremely "dumb" (because hardwired). It can be quite a lot of fun to work with these, though.
If this is a must, brutal (and safe) methods like using high-capacity de-magnetizers, sanding off the platter cannot be used.
/dev/hdX) something like below and repeat it 50 times. That should do the job.
I guess you want to re-use the hard drive after deleting? Then you are stuck with software methods. A stated before you will need MANY truly random overwrites before the disk really is not recoverable. With Linux try (for drive
dd if=/dev/random of=/dev/hdX
Even if they had a tamperproof stream from first medium to speaker: nothing is easier than setting even the stereo boxes into a closed cabinet along with a decent microphone.
(Un)Fortunately our brain still has (copyable) analogue input and especially no bit scrambling - so all and each system share this kind of weakness.
You maybe remember the videotaped movie theater movies (StarWars E1, Martix, whatever)? Even with (oldfashioned) media control (movie reels) copy protection and pay-per-view (in the theaters) it was not a problem to get fairly decent quality.
Sure? I once witnessed an Win update (don't remember which) doing the update by removing the old files (known to be old Win versions) and copying in the new files. The (formerly) used (and removed) Microsoft driver was NOT among the ones listed with the new Win. Oops.
Yes, drivers are tied to hardware - but they are also part of a software system. And if that system is too good in removing its own "garbage", there is a problem.
Then it depends how the "BuzzWordWare(tm)" company will be handling OEM rollouts - and rollout of the updates. Will the OEM hotlines be able to offer immediate help ("just keep on paying and we prolong your subscription for you") or will they have to forward that to the other company?
Another problem may become orphaned device drivers. What if the "BuzzWordWare(tm)" company decides that this or that driver is outdated (as has happened to some of the Win9x drivers)? Then all of a sudden the whole system will stop working - no chance of keeping on using the old driver. You have subscribed, so you have to update or stop using, remember?
...the date when Mir comes flaming down. Maybe they will give her a tiny boost in between to gain a bit of time for fruitless discussions. So I guess it will take a bit - let's assume april (with the boost).
2001-04-16 19:47:13
Rough guess on time - visibility estimated from parameters.
I (help desk) am on internet with address 9.9.9.9 (maybe hide-NATed, maybe not). Two users with problem are on PCs 10.10.10.10 and 10.11.11.11 (private, internal network). Both are hide-translated (N:1 NAT) to 2.2.2.2 on the internet side. So how can I connect to the first, how to the second one? Which IP address do I use for the one or other?
If you do 1:1 NAT (e.g. 2.10.10.10 and 2.11.11.11) you won't gain any security from NAT as you can still address each client individually - and thus leaving them vulnerable to DoS (flooding) attacks.
The "ASP" biztalk aside (that's just a funky workd used here for a webserver with request forms), these seem to be standard helpdesk companies which use a remote management tool (like VNC, ReachOut, pcAnywhere, CCM, etc).
You basically allow the help desk to remotely control your employee's client PCs. This means that - if you do not throw a dedicated line or big VPN solution - every single PC has to be addressable to internet. Even with a well designed firewall in place this introduces quite a risk for DoS attacks against the clients.
So if you open your company to be remotely managed by external help desks keep an eye on your security needs. Especially take care that you do not open a backdoor through the help desk system (your_net --> VPN --> help desk --> VPN --> other_net) and that liabilities and NDAs are safe and watertight for you.
Remotely managing clients can ease administration quite a bit (as posted before), but can become quite a hassle if the system does not work properly.
And make sure security does not become second. The admin tool either must require the user to confirm session shareing and enable him to immediately regain full control (by pressing a(ny) key) in case of problems - or enforce login as different user (which will log out the current user if he does not abort that with confirming a "really logout" popup box). Otherwise you will not be able to tell who did this or that possibly unpleasant change.
The concept is absolutely not new. Every FDDI ring topology could be used for/with this. Whereas Token ring only allows one single packet to circulate within the ring, this "new" concept allows, no, requires the ring to be filled up.
The "new" idea was not the ring as network topology nor the storage itself (has been done acoustically as stated above), but to coordinate client-server clusters with this. But clusters organized in ring-topology are not new either.
In the proposed topology the master server just injects "to-be-done" packets into the ring. The clients pick up (and remove) one packet from the ring each time they want to start crunching the next work packet. The other packets will be circulating the ring until solved.
Main problem is that one will be either wasting a lot of transmission capacity for idle data circulation - or be running into capacity problems. There is a reason why most detail work when designing clusters goes into designing the optimum network architecture for the specific problem...