I've made somewhat extensive use of D-Link gear, using it to provide Internet for a 600+ people at a conference, and it performed very well. We had around 20 of their $130-ish 2.4+5.2GHz APs from 3 or 4 years ago. These were not 802.11n though, this pre-dated n.
Sean
The best program to learn from is the one that you have some idea for improving, something you use and can make better. The first patch I made to open source software was for bash. At the time I was use HP-UX at work and Linux and HP-UX at home. The KSH under HP-UX would do "tab expansion" with Esc-Esc, and it was killing me going between Esc-Esc and Tab (neither worked on the other). I first tried making a macro for it, but found that I had to change the C code to make it work.
So, don't look for the prettiest code to just read -- get in the game and poke at something you use.:-)
I have run the networking at several 600-1200 attendee conferences, and have a few things you might want to try...
If any of your devices can use 5.2GHz, make sure you deploy APs for that. 5.2GHz has way more spectrum, and in my experience it tends to work where 2.4GHz is pretty spotty. Try deploying with fairly narrow beam antennas like 90 degrees, so you are just covering your booth, ideally mount it up high looking down. Run at the lowest power setting you can. Use 802.11n equipment, which often seems to have better antennas.
In the end though, 2.4GHz at conferences can be very tough... There just isn't enough spectrum there. My primary recommendation to attendees at the conference I run the wireless for is: Use 5.2GHz.
The summary makes it sound like "squeezing" 1TB into a laptop drive is impressive, but with 600GB SSDs in the same form-factor (admittedly at almost 10x the price), I'm just not overwhelmed... Especially with the recent stories about optical discs storing 500GB RSN. And the SSD is going to be able to survive being dropped without losing all that data...
And as far as performance, the summary says at 5400RPM it bests the 7200RPM competitors... That's really only true for raw streaming, say video or audio production work. People seem to be blinded by the MB/sec rate and forgetting the average access latency -- which IMHO is the most important factor in almost all cases. I had a client who was pushing back on the 15K RPM discs I recommended for their database several years ago, because the 7.2K RPM discs had a higher MB/sec number. Not for their database, they don't...
Access latency is what, in most cases, makes a computer feel slow.
My comment was, of course, fairly tongue in cheek, but this issue also seemed like a fairly amateur mistake to make. Especially for Google, who seems to in general work pretty hard to avoid making mistakes. I will agree that the Google+ roll-out has been pretty smooth for going from 10K users to 10M users in a week or two.
Especially odd is that this happened while the service was being throttled on the invitations. Seems like the first thing you'd do before opening up more invitations would be to check the capacity numbers from the previous round of invitations, and make adjustments to quotas.
Seems like there were quite a few failures:
Monitoring that didn't detect the impending resource exhaustion soon
enough for action to be taken.
The application didn't throttle itself when resources were exhausted.
Invitations were opened back up when capacity was was not there to
handle it.
Quotas were probably provisioned too tightly.
This reminds me of a conversation I was having with a google engineer
where he was expressing dismay at my mention that we had drastically
overprovissioned one of our services. He didn't seem to understand that in
my environment a few orders of magnitude overprovisioning was the minimum I
could do and still get the level of service resiliancy I needed. He also
didn't seem to "get" that I was talking about overprovisioning versus
average use (because I was making a point about that), versus potential
peak use.
Google is very interested in right-sizing their capacity, not
surprisingly. However, when you're deploying a service that's meant to
replace facebook, and your capacity planning is known to not be able to
handle new services, you probably need to think about switching out of
"rightsizing" mode and into "spending money like a drunken sailor" mode.
If this happened with the invitation system, it could have happened to
the posting or plus 1 systems as well, I imagine, and in that cause the bad
publicity from it could have meant the difference between Google+ being
taken seriously and it being a joke.
Anyway, I'm sure Google is learning from this experience. I know
I am, it's given me a great idea for modifying our monitoring to
prevent a similar problem, even for newly deployed services without much
capacity analysis history.
There's another service that is easier and more convenient than downloading DVD rips: Stop consuming their product.
You see, if you download the content off BitTorrent, the content companies see all this demand for their products and they look for a way to get revenue from that demand.
If, instead, you stop consuming their goods, they see dropping prices and offering new distribution mechanisms as increasing revenue.
So, by downloading this content, you're actually making the problems of pricing and distribution worse, for yourself and for everyone.
I like to imagine that if all the people who are downloading content off BitTorrent instead just stopped consuming their content entirely, the content providers would get a big, giant wake-up call and start paying attention to people.
One day soon, our best and brightest scientists will create a way to alert operations staff about discs reaching 90% capacity so action can be taken before they reach 100%. In that utopia, this sort of thing won't happen.
The best Linux distribution to try as a beginner is the one that your friends and family use, because they'll likely be the ones providing your tech support. Try one that has a "Live CD" so you can try it without installing it to make sure it supports the devices in your computer.
As long as their members are held to the same standards... If they abuse fair use, for example, they're required to pull their products, participate in "copyright awareness" programs, and they can only visit the top 200 websites.
But... At 8pm on a Saturday night they had a component I could use to repair the power supply in my friend's projector. The "good" place for this stuff closes at noon on Saturdays (I'm rarely ready to buy electronics before noon on Saturday), and while Radio Shack didn't have exactly what I was looking for (an individual diode), they *DID* have a full bridge rectifier, which I was able to use in place of the discrete diodes to repair it.
So, put down Radio Shack all you want, but damn it I still respect them.
There's a *VERY* easy solution to this: "No thanks." I've never had one of their clerks have an issue with me answering their address question with "No thanks". It hasn't even phased them...
When you're writing a song, how much research do you do into it it ensure accuracy? For example, in the Mandlebrot Set song (excellent analogy "Rorschach Test on fire", BTW), were you just playing the odds, or do you KNOW that it wasn't an unseasonably warm November evening for his birth? Thanks for all the great music.
Many people do not understand that "iptables-restore" is an atomic operation. In fact, the head of the Fedora Infrastructure team didn't realize this until just within the last year (we were having lunch when it came up). I always edit the "iptables-save" output, adding new rules as I need them, and then "iptables-restore" them. Usually this is via editing/etc/sysconfig/iptables and running "service iptables start" on Fedora/CentOS). However, you definitely can add and delete specific rules to running firewalls.
I've been playing with dual and big monitors lately, and my thinking on this subject is:
Do developers need dual monitors? No. Is there any reason not to provide developers dual monitors? No.
Modern IDEs use an amazing amount of real-estate. Having more screen real-estate can help prevent you from having to "change focus" to another virtual workspace, instead of completely switching context you can just flick your eyes to the IRC client, for example. But on the other hand, a small monitor doesn't have space for the other junk, so it can help me focus on a single task, if I resist the urge to check e-mail, the web, IRC (used primarily for company communications, but still often a distraction). But having lots of screen real-estate can make these interruptions much less intrusive.
After having tried it, I'd rather have one large monitor than several smaller ones. But, it's pretty cool to have multiple large monitors.:-)
Hey, thanks to those of you who actually found a point in my post instead of jumping all over the fact that I know absolutely nothing about airplanes.:-)
In WWII, the US was building up a fuel store by fueling up B52s and flying them across the Himalayas. But, depending on weather conditions, sometimes they would need to take on fuel at the depot to make the return trip. The implication in what I've read about this is that they were spending the majority of the fuel on the trip, to deposit a little fuel at the destination, like driving across the state to deliver a couple of gallons of petrol.
When google gave us a wake-up call that someone in a van could drive around and gather all sorts of information we didn't realize we were broadcasting.
A friend of mine has a theory that Anonymous is backed by the US government to help them gain support for further eroding our privacy. My initial reaction was "that's crazy", but then it reminded me of Operation Northwoods, which had the backing of every member of the Joint Chiefs of Staff.
You make some good points, that it would be hard to put a laser pointer on a jet that far away, and track it at 160MPH. However, it would be much easier if, instead of it passing by you, it were heading straight for you. Then it's clearly a case of self-defense.
Does it? Are there any hard figures to prove that?
It really depends. My feeling on it is that if you aren't going to give driving the attention it's due, you probably aren't going to whether you have those gadgets or not.
On the other hand, automatic climate control lets me set and forget the temperature and not have to adjust it nearly as much. Automatic dimming mirrors mean I don't have to be distracted with those. Stereo that plays my digital music library means I'm not messing around with changing stations. A side view mirror that automatically points down when I shift into reverse gives me better visibility. A nav system means I'm not dealing with maps and reading printed directions...
I consider all these things to be safety improvements, over all.
You know, 3 months ago I was right with you. I came quite close to buying a 4.2L v8 car with the intention of adding a supercharger to it. But the German designers of that car don't believe in many of the toys. After looking around at toys versus horsepower, I'm looking at a Japanese hybrid with: remote pre-heat/cool, radar cruise control and collision avoidance, back-up camera, lane departure warning, and all sorts of other toys not available on the sports sedan.
Of course, it's made a little easier by already having a 4.2L v8 car, so having one with the toys and one with the horsepower is probably more reasonable.:-) But this/. story really struck a chord with me based on my recent shopping.
Be sure to start this a few months before you leave./dev/urandom is INCREDIBLY slow, second only to/dev/random.../dev/frandom (which requires a custom kernel module) is probably "good enough".
I've been imagining that it might be the camel's nose under the tent. Hopefully there will be some publishers who take maximum advantage of the book loaning and see a big benefit from it. I'm not expecting the big name publishers to take advantage of it initially, but hopefully some small names will and will gain from it. Optimistic, I know, but I'm looking on the bright side.
The ZFS-FUSE setup is fantastic. For most things you are very much limited by platter speed; I've found the performance to be quite acceptable.
As far as stability goes, the 0.6.9 release, which has been out for around 3 or 4 months, has been exceptional. I did extensive stress testing of it over the last 9 months or so, and all the issues I found were resolved (quickly) by the ZFS-FUSE folks.
I currently have a 16TB backup system running with something around 2,000 snapshots and 80% space used, and it works just great. I also have a personal storage system that I have been running ZFS-FUSE on for around 3 years now, and it also has been great. I was originally running 0.5.0 on it but upgraded to the 0.6.9 after my above stress testing. It also has been great.
I used Nexenta in the past, it was ok but I think there were definite ZFS issues in it at that time, maybe 3 years ago. The systems I had would reboot every 30 to 60 days. Then I upgraded it to the latest Nexenta a year or 18 months later and had all sorts of data loss issues while trying to do the zfs send/recv from the old systems. An annoyance with Nexenta was the hardware support; coming from Linux which supports just about anything to having to dig around to find compatible storage controllers... I recently (maybe 10 months ago now, before I started really hard down the ZFS-FUSE testing path) tried OpenSolaris and ran into some weirdness where I did the install, then did an update and it spent an hour downloading updates, then bombed out. So I started it again and it spent an hour downloading updates and bombed out.
There are two reasons I'm using ZFS-FUSE so heavily: 3 years ago there was no option for encrypting my ZFS storage system in Solaris, and I just am so much more comfortable with Linux than Solaris. My home storage system stores a lot of private data, that I want to have close at hand at home, but if someone steals it I don't want to worry about the scanned check images and bills we have saved there, etc... So crypto was a huge deal for me in that server.
Slashdot Mirror
I've made somewhat extensive use of D-Link gear, using it to provide Internet for a 600+ people at a conference, and it performed very well. We had around 20 of their $130-ish 2.4+5.2GHz APs from 3 or 4 years ago. These were not 802.11n though, this pre-dated n. Sean
The best program to learn from is the one that you have some idea for improving, something you use and can make better. The first patch I made to open source software was for bash. At the time I was use HP-UX at work and Linux and HP-UX at home. The KSH under HP-UX would do "tab expansion" with Esc-Esc, and it was killing me going between Esc-Esc and Tab (neither worked on the other). I first tried making a macro for it, but found that I had to change the C code to make it work.
So, don't look for the prettiest code to just read -- get in the game and poke at something you use. :-)
I have run the networking at several 600-1200 attendee conferences, and have a few things you might want to try...
If any of your devices can use 5.2GHz, make sure you deploy APs for that. 5.2GHz has way more spectrum, and in my experience it tends to work where 2.4GHz is pretty spotty. Try deploying with fairly narrow beam antennas like 90 degrees, so you are just covering your booth, ideally mount it up high looking down. Run at the lowest power setting you can. Use 802.11n equipment, which often seems to have better antennas.
In the end though, 2.4GHz at conferences can be very tough... There just isn't enough spectrum there. My primary recommendation to attendees at the conference I run the wireless for is: Use 5.2GHz.
The summary makes it sound like "squeezing" 1TB into a laptop drive is impressive, but with 600GB SSDs in the same form-factor (admittedly at almost 10x the price), I'm just not overwhelmed... Especially with the recent stories about optical discs storing 500GB RSN. And the SSD is going to be able to survive being dropped without losing all that data...
And as far as performance, the summary says at 5400RPM it bests the 7200RPM competitors... That's really only true for raw streaming, say video or audio production work. People seem to be blinded by the MB/sec rate and forgetting the average access latency -- which IMHO is the most important factor in almost all cases. I had a client who was pushing back on the 15K RPM discs I recommended for their database several years ago, because the 7.2K RPM discs had a higher MB/sec number. Not for their database, they don't...
Access latency is what, in most cases, makes a computer feel slow.
Especially odd is that this happened while the service was being throttled on the invitations. Seems like the first thing you'd do before opening up more invitations would be to check the capacity numbers from the previous round of invitations, and make adjustments to quotas.
Seems like there were quite a few failures:
This reminds me of a conversation I was having with a google engineer where he was expressing dismay at my mention that we had drastically overprovissioned one of our services. He didn't seem to understand that in my environment a few orders of magnitude overprovisioning was the minimum I could do and still get the level of service resiliancy I needed. He also didn't seem to "get" that I was talking about overprovisioning versus average use (because I was making a point about that), versus potential peak use.
Google is very interested in right-sizing their capacity, not surprisingly. However, when you're deploying a service that's meant to replace facebook, and your capacity planning is known to not be able to handle new services, you probably need to think about switching out of "rightsizing" mode and into "spending money like a drunken sailor" mode.
If this happened with the invitation system, it could have happened to the posting or plus 1 systems as well, I imagine, and in that cause the bad publicity from it could have meant the difference between Google+ being taken seriously and it being a joke.
Anyway, I'm sure Google is learning from this experience. I know I am, it's given me a great idea for modifying our monitoring to prevent a similar problem, even for newly deployed services without much capacity analysis history.
You see, if you download the content off BitTorrent, the content companies see all this demand for their products and they look for a way to get revenue from that demand.
If, instead, you stop consuming their goods, they see dropping prices and offering new distribution mechanisms as increasing revenue.
So, by downloading this content, you're actually making the problems of pricing and distribution worse, for yourself and for everyone.
I like to imagine that if all the people who are downloading content off BitTorrent instead just stopped consuming their content entirely, the content providers would get a big, giant wake-up call and start paying attention to people.
One day soon, our best and brightest scientists will create a way to alert operations staff about discs reaching 90% capacity so action can be taken before they reach 100%. In that utopia, this sort of thing won't happen.
The best Linux distribution to try as a beginner is the one that your friends and family use, because they'll likely be the ones providing your tech support. Try one that has a "Live CD" so you can try it without installing it to make sure it supports the devices in your computer.
As long as their members are held to the same standards... If they abuse fair use, for example, they're required to pull their products, participate in "copyright awareness" programs, and they can only visit the top 200 websites.
The first thing I look for is contributions to open source software projects. But, we do open source related IT services. And it's rare to find.
But... At 8pm on a Saturday night they had a component I could use to repair the power supply in my friend's projector. The "good" place for this stuff closes at noon on Saturdays (I'm rarely ready to buy electronics before noon on Saturday), and while Radio Shack didn't have exactly what I was looking for (an individual diode), they *DID* have a full bridge rectifier, which I was able to use in place of the discrete diodes to repair it. So, put down Radio Shack all you want, but damn it I still respect them.
There's a *VERY* easy solution to this: "No thanks." I've never had one of their clerks have an issue with me answering their address question with "No thanks". It hasn't even phased them...
When you're writing a song, how much research do you do into it it ensure accuracy? For example, in the Mandlebrot Set song (excellent analogy "Rorschach Test on fire", BTW), were you just playing the odds, or do you KNOW that it wasn't an unseasonably warm November evening for his birth? Thanks for all the great music.
Many people do not understand that "iptables-restore" is an atomic operation. In fact, the head of the Fedora Infrastructure team didn't realize this until just within the last year (we were having lunch when it came up). I always edit the "iptables-save" output, adding new rules as I need them, and then "iptables-restore" them. Usually this is via editing /etc/sysconfig/iptables and running "service iptables start" on Fedora/CentOS). However, you definitely can add and delete specific rules to running firewalls.
I've been playing with dual and big monitors lately, and my thinking on this subject is: Do developers need dual monitors? No. Is there any reason not to provide developers dual monitors? No. Modern IDEs use an amazing amount of real-estate. Having more screen real-estate can help prevent you from having to "change focus" to another virtual workspace, instead of completely switching context you can just flick your eyes to the IRC client, for example. But on the other hand, a small monitor doesn't have space for the other junk, so it can help me focus on a single task, if I resist the urge to check e-mail, the web, IRC (used primarily for company communications, but still often a distraction). But having lots of screen real-estate can make these interruptions much less intrusive. After having tried it, I'd rather have one large monitor than several smaller ones. But, it's pretty cool to have multiple large monitors. :-)
Hey, thanks to those of you who actually found a point in my post instead of jumping all over the fact that I know absolutely nothing about airplanes. :-)
In WWII, the US was building up a fuel store by fueling up B52s and flying them across the Himalayas. But, depending on weather conditions, sometimes they would need to take on fuel at the depot to make the return trip. The implication in what I've read about this is that they were spending the majority of the fuel on the trip, to deposit a little fuel at the destination, like driving across the state to deliver a couple of gallons of petrol.
When google gave us a wake-up call that someone in a van could drive around and gather all sorts of information we didn't realize we were broadcasting.
A friend of mine has a theory that Anonymous is backed by the US government to help them gain support for further eroding our privacy. My initial reaction was "that's crazy", but then it reminded me of Operation Northwoods, which had the backing of every member of the Joint Chiefs of Staff.
You make some good points, that it would be hard to put a laser pointer on a jet that far away, and track it at 160MPH. However, it would be much easier if, instead of it passing by you, it were heading straight for you. Then it's clearly a case of self-defense.
Does it? Are there any hard figures to prove that?
It really depends. My feeling on it is that if you aren't going to give driving the attention it's due, you probably aren't going to whether you have those gadgets or not.
On the other hand, automatic climate control lets me set and forget the temperature and not have to adjust it nearly as much. Automatic dimming mirrors mean I don't have to be distracted with those. Stereo that plays my digital music library means I'm not messing around with changing stations. A side view mirror that automatically points down when I shift into reverse gives me better visibility. A nav system means I'm not dealing with maps and reading printed directions...
I consider all these things to be safety improvements, over all.
Sean
You know, 3 months ago I was right with you. I came quite close to buying a 4.2L v8 car with the intention of adding a supercharger to it. But the German designers of that car don't believe in many of the toys. After looking around at toys versus horsepower, I'm looking at a Japanese hybrid with: remote pre-heat/cool, radar cruise control and collision avoidance, back-up camera, lane departure warning, and all sorts of other toys not available on the sports sedan.
:-) But this /. story really struck a chord with me based on my recent shopping.
Of course, it's made a little easier by already having a 4.2L v8 car, so having one with the toys and one with the horsepower is probably more reasonable.
Be sure to start this a few months before you leave. /dev/urandom is INCREDIBLY slow, second only to /dev/random... /dev/frandom (which requires a custom kernel module) is probably "good enough".
I've been imagining that it might be the camel's nose under the tent. Hopefully there will be some publishers who take maximum advantage of the book loaning and see a big benefit from it. I'm not expecting the big name publishers to take advantage of it initially, but hopefully some small names will and will gain from it. Optimistic, I know, but I'm looking on the bright side.
Sean
The ZFS-FUSE setup is fantastic. For most things you are very much limited by platter speed; I've found the performance to be quite acceptable.
As far as stability goes, the 0.6.9 release, which has been out for around 3 or 4 months, has been exceptional. I did extensive stress testing of it over the last 9 months or so, and all the issues I found were resolved (quickly) by the ZFS-FUSE folks.
I currently have a 16TB backup system running with something around 2,000 snapshots and 80% space used, and it works just great. I also have a personal storage system that I have been running ZFS-FUSE on for around 3 years now, and it also has been great. I was originally running 0.5.0 on it but upgraded to the 0.6.9 after my above stress testing. It also has been great.
I used Nexenta in the past, it was ok but I think there were definite ZFS issues in it at that time, maybe 3 years ago. The systems I had would reboot every 30 to 60 days. Then I upgraded it to the latest Nexenta a year or 18 months later and had all sorts of data loss issues while trying to do the zfs send/recv from the old systems. An annoyance with Nexenta was the hardware support; coming from Linux which supports just about anything to having to dig around to find compatible storage controllers... I recently (maybe 10 months ago now, before I started really hard down the ZFS-FUSE testing path) tried OpenSolaris and ran into some weirdness where I did the install, then did an update and it spent an hour downloading updates, then bombed out. So I started it again and it spent an hour downloading updates and bombed out.
There are two reasons I'm using ZFS-FUSE so heavily: 3 years ago there was no option for encrypting my ZFS storage system in Solaris, and I just am so much more comfortable with Linux than Solaris. My home storage system stores a lot of private data, that I want to have close at hand at home, but if someone steals it I don't want to worry about the scanned check images and bills we have saved there, etc... So crypto was a huge deal for me in that server.
Sean