If they plan [...]
PLAN?
You think they PLAN?
HAHAHAHAHA!
Topic: We are victims of our own success - page 2. (Read 4673 times)
PLAN?
You think they PLAN?
HAHAHAHAHA!
If they plan on closing the exchange for 2-4 hours in the next 12 to 24, how about a heads up and notification on specifically when this might be. If I had to guess, they will do this around 3pm EST with no notification at the very moment the price becomes most volatile.
Clowns.
And what has Coinlab done to help the situation at Mt.Gox? When Peter Veseness announced the partnership with Mt. Gox, it was described as all kitten and rainbows. I can't figure out what advantages that partnership has brought.
Clowns.
And what has Coinlab done to help the situation at Mt.Gox? When Peter Veseness announced the partnership with Mt. Gox, it was described as all kitten and rainbows. I can't figure out what advantages that partnership has brought.
According to the IRC conversation with Magicaltux, the exchange engine and front end are running on the SAME SERVER
If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
MtGox was always PHP/mysql, including the trading engine. Do you know if it is still the same?If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
Umm, isn't using PHP at trading engine bit incompetent? For web stuff yes, but trading engine?
Haven't they really not hired any professionals? With industry experience?
MtGox was always PHP/mysql, including the trading engine. Do you know if it is still the same?
The more I learn about mtgox, the more incompetent they seem. PHP/MySQL? For a billion dollar market?
I'm revising my opinion. Negligence is a much more appropriate word than incompetence.
According to the IRC conversation with Magicaltux, the exchange engine and front end are running on the SAME SERVER
If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
MtGox was always PHP/mysql, including the trading engine. Do you know if it is still the same? If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
Yeah.
Maybe we need an even more catastrophic event at Mt.Gox before people start to rethink using it.
Wiping out a BILLION DOLLARS wasn't enough for you?
I do not use Mt.Gox. But I doubt it will be enough for the majority of users.
lol, from bitcoinity:
last trade: 4 minutes ago.
Mtgox lag: 8 seconds.
wat.
last trade: 4 minutes ago.
Mtgox lag: 8 seconds.
wat.
Yeah.
Maybe we need an even more catastrophic event at Mt.Gox before people start to rethink using it.
Wiping out a BILLION DOLLARS wasn't enough for you?
They do have DDoS protection. Cost them a pretty penny in 2011. I'm assuming they still have it. All that other stuff, yeah, you're probably right.
DDOS protection against external traffic is logical...but simultaneously they have a giant "DDOS welcome" by allowing arbitrarily small trades and bots to flood their crappy engine wihtout restrictions. Thus any registered user can cause massive disruption through an atttack via penny trades.
They have known for ages that their engine has flaws. Don't want to fix that, or it's too complicated in the near term? At least mitigate overloading scenarios by imposing a minimum trade size in BTC or USD, or raise fees to make HFT difficult, or restrict the trading API to only trusted businesses and payment proccessors, until you can get your scalability fixed.
Because there's a feedback loop:
Existing volume creates a good environment for trading, which attracts volume. It's a virtuous cycle that is only broken when they fuck up (more and more, as the volume increases).
There's a lot of 'inertia" from people who are not willing to try another exchange because of the hassle of verification and signup (weeks), wire transfers (weeks) etc.
Which is why people need to use this opportunity to STAY ANGRY and remember why they are angry. Use that anger to motivate them to sign up to new exchanges today. So that next time this happens (and it will, sooner rather than later), they can dump gox.
Existing volume creates a good environment for trading, which attracts volume. It's a virtuous cycle that is only broken when they fuck up (more and more, as the volume increases).
There's a lot of 'inertia" from people who are not willing to try another exchange because of the hassle of verification and signup (weeks), wire transfers (weeks) etc.
Which is why people need to use this opportunity to STAY ANGRY and remember why they are angry. Use that anger to motivate them to sign up to new exchanges today. So that next time this happens (and it will, sooner rather than later), they can dump gox.
Yeah.
Maybe we need an even more catastrophic event at Mt.Gox before people start to rethink using it.
Because there's a feedback loop:
Existing volume creates a good environment for trading, which attracts volume. It's a virtuous cycle that is only broken when they fuck up (more and more, as the volume increases).
There's a lot of 'inertia" from people who are not willing to try another exchange because of the hassle of verification and signup (weeks), wire transfers (weeks) etc.
Which is why people need to use this opportunity to STAY ANGRY and remember why they are angry. Use that anger to motivate them to sign up to new exchanges today. So that next time this happens (and it will, sooner rather than later), they can dump gox.
Existing volume creates a good environment for trading, which attracts volume. It's a virtuous cycle that is only broken when they fuck up (more and more, as the volume increases).
There's a lot of 'inertia" from people who are not willing to try another exchange because of the hassle of verification and signup (weeks), wire transfers (weeks) etc.
Which is why people need to use this opportunity to STAY ANGRY and remember why they are angry. Use that anger to motivate them to sign up to new exchanges today. So that next time this happens (and it will, sooner rather than later), they can dump gox.
What I do not understand from a rational perspective (well, I can understand this from a mass psychological one):
It's a free market. Why does everyone use Mt.Gox?
There are quite a few alternatives.
It's a free market. Why does everyone use Mt.Gox?
There are quite a few alternatives.
According to the IRC conversation with Magicaltux, the exchange engine and front end are running on the SAME SERVER
If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
One server? With amount of money I heard they have around?
At minimum the web services should on different servers than the trading...
Truly amateurs, and I'm a amateur...
According to the IRC conversation with Magicaltux, the exchange engine and front end are running on the SAME SERVER
If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
The worst thing is that what you've pointed out has been obvious since 2011. They've had two years between when these problems were identified and now to implement proper solutions, and either did nothing or did something which has not been effective. If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
According to the IRC conversation with Magicaltux, the exchange engine and front end are running on the SAME SERVER
If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
They do have DDoS protection. Cost them a pretty penny in 2011. I'm assuming they still have it. All that other stuff, yeah, you're probably right.
According to the IRC conversation with Magicaltux, the exchange engine and front end are running on the SAME SERVER
If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
If that is true, it points to a level of amateurish hackery which is unbelievable.
A system like this should have multiple geographically redundant load balancers, multi-tier hierarchy of workers, web engines, trade engines, market data and charting engines. The DNS should be run by an independent DNS provider and they should have DDoS protection, regional failover, graceful degradation, etc etc.
It seems they have none of that, nor any clue how to do any of it. It's not even expensive nowadays, it just requires knowledge of scalable distributed systems.
The only weakness of Bitcoin currently is just a lack of exchanges to diversify the network.
And who knows what kind of hardware they are running this on. A professional op would not have gone down like that. There should have been a failover ready to roll in such a scenario, whatever it is it is not fault tolerant in any way which isn't good.
I imagine I will begin one of my own someday, though setting up an exchange isn't all that documented yet.
And who knows what kind of hardware they are running this on. A professional op would not have gone down like that. There should have been a failover ready to roll in such a scenario, whatever it is it is not fault tolerant in any way which isn't good.
I imagine I will begin one of my own someday, though setting up an exchange isn't all that documented yet.
So basically this is not MtGox lying, this is MtGox doing the spin cycle. "We stepped in poo, we're great ecologists".
Basically this is a bunch of teenage dorks caught deer-in-headlights during an attempt to parlay everything they've learned about how to "seduce womenz" into the business of money exchanging. It's working about on par with what you'd expect.
Basically this is a bunch of teenage dorks caught deer-in-headlights during an attempt to parlay everything they've learned about how to "seduce womenz" into the business of money exchanging. It's working about on par with what you'd expect.
Is Mt.Gox considering the lag caused by a few trades per second a success?
Is is still a success if the trades are generated by bots buying and selling 0.0001 BTC back and forth?
Is it still a success if the trading bots are doing these trades with the intention of causing more lag, or is it suddenly a DOS then?
Is is still a success if the trades are generated by bots buying and selling 0.0001 BTC back and forth?
Is it still a success if the trading bots are doing these trades with the intention of causing more lag, or is it suddenly a DOS then?
YES
Is Mt.Gox considering the lag caused by a few trades per second a success?
Is it still a success if the trades are generated by bots buying and selling 0.0001 BTC back and forth?
Is it still a success if the trading bots are doing these trades with the intention of causing more lag, or is it suddenly a DOS then?
Is it still a success if the trades are generated by bots buying and selling 0.0001 BTC back and forth?
Is it still a success if the trading bots are doing these trades with the intention of causing more lag, or is it suddenly a DOS then?
Jump to: