For iptables, there is a preprocessor called Shorewall writen by Tom Eastap. Wonderful thing that is maintained pretty well. Such preprocessors makes life easy. This discussion for easier syntax via intepreters is like the classic debate between high level GUI and commandline. Learning commandline lets one appreciate the innards enough to us the utility well enough while GUI would present most common options.
Wrong. India has more regulations like the British system it is based on. Enforcement is lax. Again, all systems can be fooled and worked around, SOX or no SOX. US has had its own share of junk bonds and pyramid/Ponzi schemes where such frauds are done with elan. Does not make them any less criminal.
India socially too has a taboo against cheating - stronger than many others. Very obviously, this did not work in Satyam. I'm Indian and am angry that these guys have made a laughing stock of India in the world's eyes.
I do agree that most guys in the US do not know what clicks in India.
Mohan
I'm Indian. I think such flame baits are a bloody joke. Most Americans think they have god given or endowed knowledge to do things better just by the gift of the gab. How else do you explain Bush getting elected twice??
I've been in the US and have worked with so called successful execs. All this talk of incompetencies apply to Americans too. Like all capitalist tendencies, Americans have learned to pay less for incompetencies, thats all. And they have cost savings to justify the action. I'll bet the US can do with a fifth of the IT and still do better if they are multi-skilled and committed.
Mohan
It is difficult to convert. This is also a function of the fat that most companies use Windoze for development. I'd personally think it will help to get cos like Google give a talk on opportunities/SoC etc where open source is distinctly preferred.
I think perceptions of speed threshold that standard PC hardware can deliver need revision. I remember a thread of similar nature about a month ago. Standard Linux can saturate 2x1GB links for 1500 packet sizes with iptbles running. A Sunfire x2100 running an Opteron at 2Ghz with 2 PCI Xpress ethernet ports forwards 900Kpps (64 byte packet) while a CISCO 3845 caps at 600Kpps. Software solutions can beat the CISCO ISR platforms hands down on perofrmance and given the ingenuity of the open source forums, delivers far richer feature set than what CISCO offers.
I see a good number of posts advocating that ASICs do a better job, are more optimised, memory bandwidth limitations/latencies etc....
I've been there and done that in terms of building a commercial software router running on standard x86 platforms and customised x86 platforms. Products like the CISCO 3845 ISR reach upto 600K packets per second raw forwarding and dip drastically in performance when ACLs, QoS etc are enables. Even for 1500B packets, these routers cannot sustain line rates for ACLs on 2 gig-ethernet interfaces bi-directional.
On a standard x86 server board running on Opteron 2GHz, a software router was able to outstrip the CISCO 3845 by 70% and retain performance with ACLs thereby making it more than twice as fast. Mind you, this was on a $1000 server hardware from a well known serverv vendor.
I'd expect the OFR to do atleast 70-80% as well.
Nothing can stand greater testimony than this to the fact that x86 hardware can deliver performance.
Can. You can get PC boards with enough slots to do this. If you want to go further, you can get 1U network hardware that are std PC baords with multiple interfaces and 24 ethernet ports etc on which you can run this. Server boards have 2 ethernet interfaces on board and give you 3 PCI slots. This will be able to meet the need of a good lot of customers.
Slashdot Mirror
For iptables, there is a preprocessor called Shorewall writen by Tom Eastap. Wonderful thing that is maintained pretty well. Such preprocessors makes life easy. This discussion for easier syntax via intepreters is like the classic debate between high level GUI and commandline. Learning commandline lets one appreciate the innards enough to us the utility well enough while GUI would present most common options.
Wrong. India has more regulations like the British system it is based on. Enforcement is lax. Again, all systems can be fooled and worked around, SOX or no SOX. US has had its own share of junk bonds and pyramid/Ponzi schemes where such frauds are done with elan. Does not make them any less criminal. India socially too has a taboo against cheating - stronger than many others. Very obviously, this did not work in Satyam. I'm Indian and am angry that these guys have made a laughing stock of India in the world's eyes. I do agree that most guys in the US do not know what clicks in India. Mohan
I'm Indian. I think such flame baits are a bloody joke. Most Americans think they have god given or endowed knowledge to do things better just by the gift of the gab. How else do you explain Bush getting elected twice?? I've been in the US and have worked with so called successful execs. All this talk of incompetencies apply to Americans too. Like all capitalist tendencies, Americans have learned to pay less for incompetencies, thats all. And they have cost savings to justify the action. I'll bet the US can do with a fifth of the IT and still do better if they are multi-skilled and committed. Mohan
It is difficult to convert. This is also a function of the fat that most companies use Windoze for development. I'd personally think it will help to get cos like Google give a talk on opportunities/SoC etc where open source is distinctly preferred.
Try OpenWorkbench for project monitoring. Have you tried Compiere for inventory?
I think perceptions of speed threshold that standard PC hardware can deliver need revision. I remember a thread of similar nature about a month ago. Standard Linux can saturate 2x1GB links for 1500 packet sizes with iptbles running. A Sunfire x2100 running an Opteron at 2Ghz with 2 PCI Xpress ethernet ports forwards 900Kpps (64 byte packet) while a CISCO 3845 caps at 600Kpps. Software solutions can beat the CISCO ISR platforms hands down on perofrmance and given the ingenuity of the open source forums, delivers far richer feature set than what CISCO offers.
I see a good number of posts advocating that ASICs do a better job, are more optimised, memory bandwidth limitations/latencies etc.... I've been there and done that in terms of building a commercial software router running on standard x86 platforms and customised x86 platforms. Products like the CISCO 3845 ISR reach upto 600K packets per second raw forwarding and dip drastically in performance when ACLs, QoS etc are enables. Even for 1500B packets, these routers cannot sustain line rates for ACLs on 2 gig-ethernet interfaces bi-directional. On a standard x86 server board running on Opteron 2GHz, a software router was able to outstrip the CISCO 3845 by 70% and retain performance with ACLs thereby making it more than twice as fast. Mind you, this was on a $1000 server hardware from a well known serverv vendor. I'd expect the OFR to do atleast 70-80% as well. Nothing can stand greater testimony than this to the fact that x86 hardware can deliver performance.
Can. You can get PC boards with enough slots to do this. If you want to go further, you can get 1U network hardware that are std PC baords with multiple interfaces and 24 ethernet ports etc on which you can run this. Server boards have 2 ethernet interfaces on board and give you 3 PCI slots. This will be able to meet the need of a good lot of customers.