I have cat-5 running through my attic and it does seem very sensitive to lightning. Don't think I've had any direct hits but at least 2 instances where there seems to have been a surge in one of the runs. Both cases resulted in burnt out ports on both the device and the switch, but no further damage. I have no idea what I can (easily) do about it.
There are configuration based approaches which work for some specific classes of attack. For the more general case the only thing you can really do is increase your downstream capacity, either by actually having fatter pipes into your DC and the requisite routers/firewalls/proxies to handle the load, or by making use of an upstream filter like cloudflare. Or both:)
I never said it was impossible:) I'm glad EFS is coming along, it's certainly a welcome addition. But, the fact it's still in preview - many years after S3 become commonplace, and it costs more than 10x as much as S3 - tells me it's not easy.
What I was trying to get across was that the reason for the popularity of the object storage model is that it benefits the storage provider - not the client.
It allows the implementation to make a lot of assumptions & simplifications, which in turn makes things like S3 possible. There's no way you could practically offer POSIX style FS access in a cloud-like environment.
Tesla has about the same number of employees and they design, build, sell and service futuristic electric cars that accelerate faster than a Formula One racing car.
First off, no they don't. The Tesla has a 0-60 of 2.8s, an F1 car does 2s or lower. You may be confusing "an F1 race car" with "the McLaren F1" which is a production road car and which Tesla stated they wanted to equal in performance.
To your main point, organizations tend to grow non-linearly in my experience. As you add more of your "primary" employee types, you need more and more support staff. As you get more clients you need more sales people to support them, more finance people to handle the money, and so on. It explodes. As well as that natural growth a company like Facebook has a huge pressure to be innovating and coming up withe next $billion idea - and with the amount of cash they have throwing people at the problem is easy. So a lot of those people are probably working on secret stuff that may never see the light of day and may only be tangentially related to their core business - a lot like Google and to a lesser extent Apple/Microsoft.
I get that prompting for a CC# at boot is not particularly friendly, but I fail to see how it is even related to being "a home console for everyone, freedom from the big publishers and for everyone to develop". Fact is very few (any?) big publishers put anything on the platform, there were tons of indie games, and you certainly can turn any box into a dev kit. Seems like they fulfilled those promises.
The fact that it was underpowered and the games didn't appeal to me is what made me sell mine (at a profit, somehow!). I don't believe anyone was screwed.
The first correct question is why on earth would someone need to access half a petabyte? In most cases the commonly accessed data is less than 1%. That's the amount of data that realistically needs to reside on disk. It never is more than 10% on such a large dataset.
Never say never. We have data sets several times larger than that which are 100% always online due to client access patterns. Not only online, but extremely latency critical. And I personally could name a dozen other companies with similar requirements.
Where I work we deal with data sets of a similar order. However, different data sets are stored differently depending on need. For online relational data where performance is critical, it's in master/slave/backup DB clusters running with 4.8TB PCIe SSDs. The backups are taken from a slave node and stored locally, plus they're pushed offsite. No tape, if we need a restore we can't really wait that long.
For data we can afford to access more slowly we use large HDFS clusters with regular SATA discs. There's a level of redundancy built in there, and where data is important enough to need a real backup (much of it is not) it is also pushed offsite. The HDFS approach has the advantage of presenting as a very large filesystem, and obviously if you're running hadoop against it there's an automatic advantage.
The problem with decrying BYOD as being "only for convenience" is that, when it comes down to it, basically all enterprise tech is "only for convenience". Tech exists within an organization to allow their employees to be more effective, more efficient, react faster, etc. That's what it's for. Convenience isn't a reason to ignore a technology, it should be the most important reason to adopt it.
I've worked in security in one of the most paranoid companies around and I totally get the need to protect the network - but the approach of just default denying everything because it's easier than figuring out how to allow something in a safe manner is lazy, and dare I say it, just for your convenience.
Huh? The speed limits on UK streets are broadly the same or higher compared to those in the US. Having driven for many years on both I really don't see much difference other than US streets are typically wider and the highways are considerably slower. I drive 30-40 on typical (sub)urban streets in both places.
There's no enforcement of jaywalking laws in plenty of the US too (e.g. NYC). It's not about safety (to my mind) but about indicating whether the car or the person has priority in that city. The UK and NYC both have large pedestrian populations which other parts of the US do not - those tend to be where jaywalking is frowned upon.
You shouldn't have to worry about backups at work because they should be handling that for you. Usually they either backup individual machines or back up shared dirs on servers and ask you to keep your stuff there. Both reasonable approaches. Running your own backup is not - it's not your data and you don't get to say where it's kept.
At home, offsite is easy, and can be done for free. I run crashplan and push about 4TB to them from my own machines - my servers also act as offsite backups for other family members who live elsewhere with less storage requirements (my mother in law has 100GB or so backed up to me for example). That peer to peer backup is free and so much better that the USB drive you sometimes remember to update once a week or so and lives in the trunk of your car. Also makes multiple backups easy - my stuff is in several places for example - single points of failure are bad.
As someone else mentioned, should your house burn down, you likely have larger concerns anyway. I disagree. Access to important data is likely to be a pretty major concern in such an incident.
The DSLReports forums have special sections for some ISPs where you can talk directly to a senior tech. I was able to get a faulty router replaced super easily there, no phone call required.
As someone who works on such large scale systems, I disagree. When you need to deal with extremely high concurrency the functional paradigm with immutable structures is a really nice way to reason about problems - I'd say it contributes significantly to reliability over standard threaded imperative code.
"Fast" is such a vague term as to be meaningless - but I can say that we typically hit the performance limits of something external (network, disk, DB) before the fact that we're in a JVM makes any difference. If your problem is purely compute then maybe it's worth looking at C or golang, but the vast majority of stuff I work on is network services, and compute is not the bottleneck.
I do 100% agree with the strongly typed bit though:)
You're saying that this mythical average iTunes user (remember everyone who has ever owned an Apple device has an account, even if they never use it) who spends $15 a year is somehow equivalent to Spotify Premium subscriber, who clearly cares enough about music to subscribe to a service and not just use the free version? What you need to find out is how much the average Premium Spotify subscriber used to spend per month on music. I'm pretty confident that'd be over $10.
Sure, that works for some companies. In the specific case that I was describing, the vast majority of employees live in the city (many within walking distance of the office) and I think they'd be unwilling to leave. I personally live in the suburbs and commute in, but I'm the exception. Looking at the companies around where I live, there are very few I'd be interested in working for.
Slashdot Mirror
I have cat-5 running through my attic and it does seem very sensitive to lightning. Don't think I've had any direct hits but at least 2 instances where there seems to have been a surge in one of the runs. Both cases resulted in burnt out ports on both the device and the switch, but no further damage. I have no idea what I can (easily) do about it.
Except there is no universal device ID on iOS. Each app sees it's own device ID, so an app which grabbed it and leaked it would benefit no one.
There are configuration based approaches which work for some specific classes of attack. For the more general case the only thing you can really do is increase your downstream capacity, either by actually having fatter pipes into your DC and the requisite routers/firewalls/proxies to handle the load, or by making use of an upstream filter like cloudflare. Or both :)
I see that one of their main highlighted features is support for easily creating Photo CDs! I've been waiting years for this - where do I sign?
I never said it was impossible :) I'm glad EFS is coming along, it's certainly a welcome addition. But, the fact it's still in preview - many years after S3 become commonplace, and it costs more than 10x as much as S3 - tells me it's not easy.
What I was trying to get across was that the reason for the popularity of the object storage model is that it benefits the storage provider - not the client.
I apologize if I misrepresented your efforts!
I'd say NFS is the evidence that's it's not practically possible :)
It allows the implementation to make a lot of assumptions & simplifications, which in turn makes things like S3 possible. There's no way you could practically offer POSIX style FS access in a cloud-like environment.
First off, no they don't. The Tesla has a 0-60 of 2.8s, an F1 car does 2s or lower. You may be confusing "an F1 race car" with "the McLaren F1" which is a production road car and which Tesla stated they wanted to equal in performance.
To your main point, organizations tend to grow non-linearly in my experience. As you add more of your "primary" employee types, you need more and more support staff. As you get more clients you need more sales people to support them, more finance people to handle the money, and so on. It explodes. As well as that natural growth a company like Facebook has a huge pressure to be innovating and coming up withe next $billion idea - and with the amount of cash they have throwing people at the problem is easy. So a lot of those people are probably working on secret stuff that may never see the light of day and may only be tangentially related to their core business - a lot like Google and to a lesser extent Apple/Microsoft.
I know I'm going to regret asking this, but how exactly is the open coc racist?
No one is taking anything away, relax. The upgrade is optional.
I get that prompting for a CC# at boot is not particularly friendly, but I fail to see how it is even related to being "a home console for everyone, freedom from the big publishers and for everyone to develop". Fact is very few (any?) big publishers put anything on the platform, there were tons of indie games, and you certainly can turn any box into a dev kit. Seems like they fulfilled those promises.
The fact that it was underpowered and the games didn't appeal to me is what made me sell mine (at a profit, somehow!). I don't believe anyone was screwed.
Never say never. We have data sets several times larger than that which are 100% always online due to client access patterns. Not only online, but extremely latency critical. And I personally could name a dozen other companies with similar requirements.
Where I work we deal with data sets of a similar order. However, different data sets are stored differently depending on need. For online relational data where performance is critical, it's in master/slave/backup DB clusters running with 4.8TB PCIe SSDs. The backups are taken from a slave node and stored locally, plus they're pushed offsite. No tape, if we need a restore we can't really wait that long.
For data we can afford to access more slowly we use large HDFS clusters with regular SATA discs. There's a level of redundancy built in there, and where data is important enough to need a real backup (much of it is not) it is also pushed offsite. The HDFS approach has the advantage of presenting as a very large filesystem, and obviously if you're running hadoop against it there's an automatic advantage.
The problem with decrying BYOD as being "only for convenience" is that, when it comes down to it, basically all enterprise tech is "only for convenience". Tech exists within an organization to allow their employees to be more effective, more efficient, react faster, etc. That's what it's for. Convenience isn't a reason to ignore a technology, it should be the most important reason to adopt it.
I've worked in security in one of the most paranoid companies around and I totally get the need to protect the network - but the approach of just default denying everything because it's easier than figuring out how to allow something in a safe manner is lazy, and dare I say it, just for your convenience.
Huh? The speed limits on UK streets are broadly the same or higher compared to those in the US. Having driven for many years on both I really don't see much difference other than US streets are typically wider and the highways are considerably slower. I drive 30-40 on typical (sub)urban streets in both places.
There's no enforcement of jaywalking laws in plenty of the US too (e.g. NYC). It's not about safety (to my mind) but about indicating whether the car or the person has priority in that city. The UK and NYC both have large pedestrian populations which other parts of the US do not - those tend to be where jaywalking is frowned upon.
Do they have videos of Greek Nationalists paying their taxes? Oh wait...no such thing exists :)
You shouldn't have to worry about backups at work because they should be handling that for you. Usually they either backup individual machines or back up shared dirs on servers and ask you to keep your stuff there. Both reasonable approaches. Running your own backup is not - it's not your data and you don't get to say where it's kept.
At home, offsite is easy, and can be done for free. I run crashplan and push about 4TB to them from my own machines - my servers also act as offsite backups for other family members who live elsewhere with less storage requirements (my mother in law has 100GB or so backed up to me for example). That peer to peer backup is free and so much better that the USB drive you sometimes remember to update once a week or so and lives in the trunk of your car. Also makes multiple backups easy - my stuff is in several places for example - single points of failure are bad.
As someone else mentioned, should your house burn down, you likely have larger concerns anyway.
I disagree. Access to important data is likely to be a pretty major concern in such an incident.
3 billion searches a year works out to 95/second by my estimations. Which is laughably insignificant. Why on earth would Apple be interested in that?
The DSLReports forums have special sections for some ISPs where you can talk directly to a senior tech. I was able to get a faulty router replaced super easily there, no phone call required.
As someone who works on such large scale systems, I disagree. When you need to deal with extremely high concurrency the functional paradigm with immutable structures is a really nice way to reason about problems - I'd say it contributes significantly to reliability over standard threaded imperative code.
"Fast" is such a vague term as to be meaningless - but I can say that we typically hit the performance limits of something external (network, disk, DB) before the fact that we're in a JVM makes any difference. If your problem is purely compute then maybe it's worth looking at C or golang, but the vast majority of stuff I work on is network services, and compute is not the bottleneck.
I do 100% agree with the strongly typed bit though :)
What bias? A bias against assuming all programmers are men? Guilty.
How does using "she" differ from using "he"? Why is "he" an acceptable default but "she" not?
You're saying that this mythical average iTunes user (remember everyone who has ever owned an Apple device has an account, even if they never use it) who spends $15 a year is somehow equivalent to Spotify Premium subscriber, who clearly cares enough about music to subscribe to a service and not just use the free version? What you need to find out is how much the average Premium Spotify subscriber used to spend per month on music. I'm pretty confident that'd be over $10.
Speak for yourself. It's not hard to become a shareholder.
Back before I switched to Spotify I'd easily spend $40 a month on music (i.e. 3 or 4 albums), so yes, in my case. YMMV of course.
Sure, that works for some companies. In the specific case that I was describing, the vast majority of employees live in the city (many within walking distance of the office) and I think they'd be unwilling to leave. I personally live in the suburbs and commute in, but I'm the exception. Looking at the companies around where I live, there are very few I'd be interested in working for.