SERVER CRASHES AGAIN - Google Cash Detective!

[richfit]

Not another server crash on another launch day... who else is annoyed by that? Well, I've been getting calls all day about this problem so I just want to come out and say, "The server is back up my friends!"

Do you think it was used to create more desire and anticipation?

What do you think about these server crashes?

Best,

Bryan

[Sean A McAlister]

It was legit...........and they happen

[Black Hat Cat]

Originally Posted by richfit

Not another server crash on another launch day... who else is annoyed by that?

What do you think about these server crashes? Do you think it was used to create more desire and anticipation?

I think you answered your own question.

[Chris Kent]

Of course it was fake, this happens all the time.

I mean, if they were all real then the no.1 priority of any launch would be "check server will handle traffic on day 1 without falling over".

Nobody buys this BS anymore.

[Trader54]

Happens all the time, gives reason to keep emailing.

It's up .. It's down ... It's back up again.

Thanks for being patient, just for that I will be adding a special bonus ... just for YOU!

[jasonl70]

Originally Posted by richfit

Do you think it was used to create more desire and anticipation?

I think it's a huge myth.. I can't see anyone stoking up the fires for a few weeks, then just snuffing it out in the middle of a huge rush sales.. I've got Jeff Walkers PLF, and he certainly never taught this (I've heard people make this accusation in the past).

I think the truth is that these bozo's who have crashes do a half-assed job of setting up their infrastructure.

Well, maybe some of the newer guys were under the impression that guru's server crashes were orchestrated, so they might be faking it now, thinking it's all part of a successfull 'guru-size' launch?

[jasonl70]

turning off a server in the middle of sales, just for an excuse to send another email, would be assinine in my opinion.

[LegitBlogger]

Who knows, maybe the crashes really happen... but
I wonder why CNN, BBC, Wikipedia, YouPorn and the
other top sites with MASSIVE traffic don't experience
such crashes. Or do they?

[Sean A McAlister]

Originally Posted by ProductCreator

Of course it was fake, this happens all the time.

I mean, if they were all real then the no.1 priority of any launch would be "check server will handle traffic on day 1 without falling over".

Nobody buys this BS anymore.

Originally Posted by Trader54

Happens all the time, gives reason to keep emailing.

It's up .. It's down ... It's back up again.

Thanks for being patient, just for that I will be adding a special bonus ... just for YOU!

Really....ITS ALL FAKE............well considering I have over 60 affiliates promoting this product and cannot get into the back office right now..... and have been speaking with Chris....behind the scenes.......because his SERVER WENT DOWN...and now has to manually enter stuff....its all made up.

This is funny because I just did a webinar last night about this kind of stuff. I was the JV Manager for a Launch that had a "Script Issue".....which was real because I was the guy pulling a 26 hour shift to get it back up. Still had to move the launch out....which caused a fiasco

People think this stuff is peaches and cream and that the process and execution of a major launch just happens. Let me tell you something....there is A LOT OF WORK that goes into a launch. There are a lot of integrations, coding, scripting etc.....AND YES....**** happens.

Let me see all of the conspiracy theorist pull off these kinds of launches without a hitch.

[dsmpublishing]

i remember the day i did my very 1st launch for new subscribers aweber crashed i was mortified but unfortunately it does happen!

[richfit]

Originally Posted by Trader54

Happens all the time, gives reason to keep emailing.

It's up .. It's down ... It's back up again.

Thanks for being patient, just for that I will be adding a special bonus ... just for YOU!

Hahahahaha that's hilarious! Yea but they are incredibly annoying!

[richfit]

Originally Posted by The IM Reporter

Really....ITS ALL FAKE............well considering I have over 60 affiliates promoting this product and cannot get into the back office right now..... and have been speaking with Chris....behind the scenes.......because his SERVER WENT DOWN...and now has to manually enter stuff....its all made up.

This is funny because I just did a webinar last night about this kind of stuff. I was the JV Manager for a Launch that had a "Script Issue".....which was real because I was the guy pulling a 26 hour shift to get it back up. Still had to move the launch out....which caused a fiasco

People think this stuff is peaches and cream and that the process and execution of a major launch just happens. Let me tell you something....there is A LOT OF WORK that goes into a launch. There are a lot of integrations, coding, scripting etc.....AND YES....**** happens.

Let me see all of the conspiracy theorist pull off these kinds of launch without a hitch.

Your Absolutely RIGHT! I agree with you. **** does happen. When you have thousands of people clicking on the same link at the same time that's what happens.

What you need to have happen is have 4-5 different purchase pages, which would be annoying but would assure that this wouldn't happen!

[JayXtreme]

I'm with Sean here...

How in bejeezes can you peeps make assumptions like that..

Picture this:

You mail a promo to your lists... but, instead of the proper link, you put a foul link in the mail on purpose... do you REALLY think that another mail is gonna bring extra results... I mean, really... do you believe that people will wait, cap in hand for you next mail?...

Nah, I don't think so...

Most people won't have time to wait around for the doors to open... you MIGHT get the hard core crowd waitin around, but tons of people will just write it off.

Jus Sayin

Jay

[Noel2010]

It's either fake or they want to spend as little as possible on servers.

A few thousands of people looking at a long sales page and the servers can't handle it??!!

BS!

[richfit]

Originally Posted by Noel2010

It's either fake or they want to spend as little as possible on servers.

A few thousands of people looking at a long sales page and the servers can't handle it??!!

BS!

It's NOT about the amount of people on the site. It's about the # of people clicking the same link at the same time. When you have over 10,000 people clicking the link and clicking the link over and over again till they get what they want, that's when you have problems.

[Mike Hill]

There are major issues today with a certain hosting company. I'm having huge problems right now as well (probably due to that dudes launch)

Mike Hill

[Alan Petersen]

Amazing to see folks clamoring to pick up a $997 plus $97/month product in droves to bring down the server during a recession.

That being said. I'm sure it's not planned but it is poor planning. Seems like every big hyped launch brings the server down. Must be cheaper/easier to deal with the down server than put the resources into figuring out how deal with scaling beyond a few hundred/thousand folks hitting the order button at the same time.

[Josh Anderson]

I received 5 emails in one day from someone promoting the GCD who kept messing up on their link.

5 emails in one day?

I think they have been drinking too much red bull or coffee or something.

I bet those five emails cost them more in lost lifetime profits for the people who got annoyed and unsubscribed than it made them in commissions.

And these were not novices... but rather owners of mega lists from the last huge launch.

---

As for server crashes...

The reason it happens so often is because the guys running these promos are not webmasters.

They many times overlook limitations on network and uplink speed on their servers.

We have one publisher in our network who manages a site for a person who is essentially a celebrity in his niche (not IM). They are launching a new membership and affiliate program with some potential large scale affiliate promotions.

I asked him about his hosting...

The answer he gave me was that they are on a Godaddy economy hosting plan.

When I told him that Godaddy artificially limits those plans to only 50 simultaneous hits at a time and blocks the 51st he had no clue.

So they upgraded... at Godaddy... to a plan that artficially limites them to only 100 simultaneous hits at a time!!!

Most launches you see are run on servers with 100mb connections. Some of these sites are laden with self hosted videos, large images, video players that auto load which multiple players on the site etc.

Get a pile of visitors all at once and presto you got a log jam.

Some of these sites are running heavy scripts on the front end and systems and databases on the back end that could potentially cause bottle necks.

Bottle necks are hard to predict sometimes until its happening. You have to be ready to move fast with database management strategies and other modifications to eliminate the bottle neck. This may not be feasible to do when the launch is on.

Also most people just are not willing to lay out the cash to get a gigabit connection and load balancing on a private rack of servers.

Luckily there are alternatives available these days with cloud server solutions.

It's hard to predict what will happen when the crush is on.

You can prepare all you want but many times you just cannot predict where the bottle neck will be.

Of course a majority of times the crashes are due to a lack of preparation and knowledge pertaining to hosting resources...

But it could be things like human error at the data center, hardware failure, hurricanes, ddos attack, hackers, and aliens from Mars.

Aliens from Mars is just about as likely as it being staged as some deluded conspiracy theorists claim. Server crash is an embarrassing, unprofitable, and undesirable event. No successful marketer stages them.

[Mike Hill]

Originally Posted by Josh Anderson

I received 5 emails in one day from someone promoting the GCD who kept messing up on their link.

5 emails in one day?

I think they have been drinking too much red bull or coffee or something.

I bet those five emails cost them more in lost lifetime profits for the people who got annoyed and unsubscribed than it made them in commissions.

And these were not novices... but rather owners of mega lists from the last huge launch.

---

As for server crashes...

The reason it happens so often is because the guys running these promos are not webmasters.

They many times overlook limitations on network and uplink speed on their servers.

We have one publisher in our network who manages a site for a person who is essentially a celebrity in his niche (not IM). They are launching a new membership and affiliate program with some potential large scale affiliate promotions.

I asked him about his hosting...

The answer he gave me was that they are on a Godaddy economy hosting plan.

When I told him that Godaddy artificially limits those plans to only 50 simultaneous hits at a time and blocks the 51st he had no clue.

So they upgraded... at Godaddy... to a plan that artficially limites them to only 100 simultaneous hits at a time!!!

Most launches you see are run on servers with 100mb connections. Some of these sites are laden with self hosted videos, large images, video players that auto load which multiple players on the site etc.

Get a pile of visitors all at once and presto you got a log jam.

Some of these sites are running heavy scripts on the front end and systems and databases on the back end that could potentially cause bottle necks.

Bottle necks are hard to predict sometimes until its happening. You have to be ready to move fast with database management strategies and other modifications to eliminate the bottle neck. This may not be feasible to do when the launch is on.

Also most people just are not willing to lay out the cash to get a gigabit connection and load balancing on a private rack of servers.

Luckily there are alternatives available these days with cloud server solutions.

It's hard to predict what will happen when the crush is on.

You can prepare all you want but many times you just cannot predict where the bottle neck will be.

Of course a majority of times the crashes are due to a lack of preparation and knowledge pertaining to hosting resources...

But it could be things like human error at the data center, hardware failure, hurricanes, ddos attack, hackers, and aliens from Mars.

Aliens from Mars is just about as likely as it being staged as some deluded conspiracy theorists claim. Server crash is an embarrassing, unprofitable, and undesirable event. No successful marketer stages them.

Nice explanation... I just learned what to look for before my big launch next month. Thanks for taking the time to cover those things.

I am one of those marketers who is clueless when it comes to all this stuff...

Very enlightening indeed!

Mike Hill

PS. Seems like more people need to hire out this sort of thing just to make sure everything goes without a hitch. (Are you for hire for this sort of thing on a launch Josh?)

[Josh Anderson]

Very enlightening indeed!

Mike Hill

PS. Seems like more people need to hire out this sort of thing just to make sure everything goes without a hitch. (Are you for hire for this sort of thing on a launch Josh?)

I do consult on hosting, video optimization, content delivery, business automation, and development strategies.

Frankly... the price of my hourly consulting fee would be a bargain compared to the cost of having a server crash. Also not long ago I laid out in the forum how one guru could cut his bandwidth bill from $25k to $7k and boost his response rate at the same time ;-)

[JonathanBoettcher]

Anyone else out there get burned in the LAST GCD launch a year or two ago?

Just wondering.

[ChasHicks]

In a past life I use to do load testing on platforms similar to what is now called "cloud computing" on large systems. When we had large releases of products to tens of thousands of users who would be using it concurrently we actually built test beds of application servers and would model the response of systems to find the bottlenecks.

Though a combination of hardware and software resources, including multiple database servers, front end servers, etc. you can be pretty well assured that the actual launch of something like this will go smoothly. It is an expensive process taking months of planning and lots of expensive human resources.

But as mentioned above, most launches of this type will not spend the money to do this level of testing and build the suitable architecture. It's not worth it when you can push everyone back a day or so and say "try later".

When the demand is great enough as is in these big launches, then you can do that.

If it was a mission critical application that HAD to work or a company would be out of business then they would take it more seriously.

Are some launch server crashes orchestrated? Maybe - maybe not. If I ever found out they were I'd black list them and expose them in a heartbeat.

After seeing lots of large launches like this I would think they'd learn a few lessons and devise a way to spread the traffic out over time. Some have done this by having early notification lists, etc.

I'm sure it is very frustrating for Chris to see these problems with the GCD launch after having spent so much time and money developing the system and getting to the launch time.

Charlie

[Mike Merz]

Sean is right ... I've been going back and forth with Chris for weeks helping him tweak the JV end of the process, and with all the launches going down ... and the Top performers booked more than a month in advance ... you may only get 1 mailing, if you're lucky ... from each major partner.

It's common that even the best converting sales process will drop 50% after 24 hours ... and 50%, again, after 48.

... why would you want to chance losing those opportunities to apply a supposed 'tactic'? ... and, potentially, lose JV partners?!?

I went out for a few hours to grab lunch ... and came back to see (as Sean did) an error message when I tried to log in. I got a hold of Chris ... who had been working with his team to correct things, and he said they were back Online ... and he sounded relieved more than anything else.

Best,

Mike Merz

[SiteSeller52]

Well, I am a JV for the product and I have
good numbers from the launch, but I was
somewhat let down by the fact that we
(the jv's) were lied to about getting access
to GCD before launch.

When I sent an email inquiring about it,
I got no response. I also never got my
first jv email, which was the subject
of my inquiry, so I had to set up a new
account.

If the person selling a new product
lies to you before a launch, do you
still promote his product and hope
he pays you? (thats what I did hope
I get paid)

[Josh Anderson]

To continue what I was mentioning above about hidden artificial limits some hosts put on their services:

The key is that you would not know that a host has artificial limits unless you actually asked because most hosts do not disclose those things. You need to ask a series of questions to any host you are considering that most people do not know to ask because there are other artificial limits that hosts can use like limits mysql queries, daily bandwidth limits (even though they advertise a monthly allotment) and more.

---

Most launches you see are run on servers with 100mb connections. Some of these sites are laden with self hosted videos, large images, video players that auto load which multiple players on the site etc.

In addition to that statement I made above even most who have dedicated servers are still on a 100mb uplink and network. If you are serving a 250kbps video and its hosted on the same server as your site you are using in the launch and that video is set to auto load when visitors landed on your site you would only be able to handle approximately 400 visitors at a time or less. Most publishers do not know how to optimize their videos and even recently one major launch used a 1000kbs video. If you were to host that on a 100mb connection following the same scenario above you would only be able to have 100 visitors at a time.

This is why hosting your video on s3 during peak demands is a better idea. So is optimizing your video bit rate in order to reduce your bandwidth bill.

If a server crash happens and it is truly due to traffic demand consuming all the bandwidth available or all the processor power available and they get the site back online it is not because the publisher made any significant changes (server hardware changes are usually very hard to make fast and can take days) but rather it is because while the server was crashed huge amounts of traffic was pouring in and once it stopped pouring in the site was accessible again.

That means that while it was offline the publisher lost out on significant potential sales. Which is why its a nightmare... not a desirable strategy that anyone would seek.

[Steven Wagenheim]

If I ever do a major launch (not likely anytime soon) the last thing I want is
a server crash, tactic or otherwise.

In my opinion, it only kills business, not helps it.

And since I know beans about tech stuff, I'm going to make sure that I at
least get some intelligent people to help me with the tech end of it.

Planned server outages?

Sorry, but I'm not buying it.

[jasonl70]

If these guys are making as much money as they do, they need to get a fat pipe to the net, and set up a server farm.

Heck, I'm getting ready to launch a service for 'bricks & mortar' companies, and there's a chance it could get big - fast. I went from thinking I needed 1 dedicated server initially, to buying 10 servers and redesigning my entire application architecture to handle quickly scaling up (and adding more servers)..

I was the tech lead several years ago for an online 'time card' project for US Air Force. Imagine 100,000 users trying to log on and 'punch in' at exactly 8am Before they brought me on, they hadn't even considered what this meant.

I was once asked to make a system for Kmart to simply log the status of updates that were sent out to windows machines in the stores. I gave them an estimate which they did not like, and one of their programmers (actually a contractor who had been there for years as unix sys admin and perl programmer) laughed at me and said he'd do it himself in a day.

It took him a few days, and he put his code onto a production server with no load testing on a friday.

Monday morning, every pc and cash register in every kmart store in the US had crashed. At which point, they gave me the go ahead to do it 'right'.

My point is, people need to plan for this stuff IN ADVANCE. And hire professional's if needed.

[Josh Anderson]

Originally Posted by ChasHicks

In a past life I use to do load testing on platforms similar to what is now called "cloud computing" on large systems. When we had large releases of products to tens of thousands of users who would be using it concurrently we actually built test beds of application servers and would model the response of systems to find the bottlenecks.

Though a combination of hardware and software resources, including multiple database servers, front end servers, etc. you can be pretty well assured that the actual launch of something like this will go smoothly. It is an expensive process taking months of planning and lots of expensive human resources.

But as mentioned above, most launches of this type will not spend the money to do this level of testing and build the suitable architecture. It's not worth it when you can push everyone back a day or so and say "try later".

When the demand is great enough as is in these big launches, then you can do that.

If it was a mission critical application that HAD to work or a company would be out of business then they would take it more seriously.

This is what every major marketer who is planning a "million dollar" launch should take into consideration.

I was once asked to make a system for Kmart to simply log the status of updates that were sent out to windows machines in the stores. I gave them an estimate which they did not like, and one of their programmers (actually a contractor who had been there for years as unix sys admin and perl programmer) laughed at me and said he'd do it himself in a day.

It took him a few days, and he put his code onto a production server with no load testing on a Friday.

Monday morning, every pc and cash register in every kmart store in the US had crashed. At which point, they gave me the go ahead to do it 'right'.

My point is, people need to plan for this stuff IN ADVANCE. And hire professional's if needed.

I love that example. Too many people take the "KMart approach" and end up paying the price.

[jasonl70]

Originally Posted by Mike Merz

Sean is right ... I've been going back and forth with Chris for weeks helping him tweak the JV end of the process, and with all the launches going down ... and the Top performers booked more than a month in advance ... you may only get 1 mailing, if you're lucky ... from each major partner.

wow.. think about that one for a minute... people so busy, they need a month notice just to send an email! Does that sound like the care free "IM" lifestyle a lot of people are looking for? the big dogs are working their tails off ..

[IMChick]

Sheesh.

I need to find a place that can explain exactly what I need to do and the how and why of it in non-techie terms that I can understand and implement and more important, can supervise in my business. I've been burned with the dp-type kmart guy a few times, and won't do that again if I can learn enough to avoid it.

The 'servers going down' during a launch is one of my recurring nightmares. I have absolutely no idea how to instruct a purchased consultant or on-site employee about this, because I myself don't know what is supposed to happen (other than you click, we get paid). Hence, the screaming nightmares.

Any ideas or sources would be great.

[templarjustice]

The funny thing is no one ever really talks about the cost of all this. Part of the reason why their sites are crashing is because even the large gurus don't want to pay he costs. To do load balancing & colocation for large launches will cost you around $2,000 to $3,500 per month. These are servers where you can pull out a drive & you still have 3 other servers running completely fine. So if one fails the whole system is still running. But you are looking at a cost of at least $2,000 per month. I personally don't see why it should even cost that much, & I believe there is a monopoly on the little guys who are basically putting in their $9.95 or $30, & $70 per month in and they still get problems. I think there is technology good enough out there to perfect everyone at a lower cost & it doesnt' take cloud computing to do this.

[JayInOrlando]

OMG. Guys in this day and age... really.

It's not hard. Upload your video files to one server, have your main site hosted on another server. Have the backed on yet another server. Use cloud hosting, use multiple servers with load balancing. If your launching a million dollar product, use a freaking million dollar hosting system. Heck even have a backup system in place in case everything else does go down.

Jay

[Matrix9162]

It's a painful day, but just be patient. The launch is still open for those who are trying to get in. I thought it would close in a few hours, but guess not. I wonder what the limit is.

[entrepenerd]

I'm still completely amazed that I haven't heard anyone talking about doing a launch using Amazon's EC2 and CloudFront service. This combined with S3 for content delivery provides immediate and flexible scaling with rollover support.

I work for a company that handles the web application for the Parks District for a major metropolitan area. 4 times a year their application receives thousands of orders per second for about a 30 minute period. As soon as the doors open for registration their servers get pounded. BUT, we pull it off without a hitch every time.

The resources are available to handle a major launch like this without missing a beat, and with the new Amazon hosting services available it's not even expensive.

With Amazon EC2 you can setup a system that fires up new servers on-demand and adds them to a load balancing cluster and when the load decreases they shut down so you're not paying for the extra resources. It's killer stuff.

I guess I'll just have to put together a how-to guide and sell it to the million dollar launchers out there

[Gail Sober]

Next time use a company that is used to accomodating adult websites then move things over to your normal server once the launch part is over.

They are used to huge amounts of traffic and have a lot of resources set aside to handle spikes.

[jasonl70]

Originally Posted by entrepenerd

I'm still completely amazed that I haven't heard anyone talking about doing a launch using Amazon's EC2 and CloudFront service. This combined with S3 for content delivery provides immediate and flexible scaling with rollover support.

I work for a company that handles the web application for the Parks District for a major metropolitan area. 4 times a year their application receives thousands of orders per second for about a 30 minute period. As soon as the doors open for registration their servers get pounded. BUT, we pull it off without a hitch every time.

The resources are available to handle a major launch like this without missing a beat, and with the new Amazon hosting services available it's not even expensive.

With Amazon EC2 you can setup a system that fires up new servers on-demand and adds them to a load balancing cluster and when the load decreases they shut down so you're not paying for the extra resources. It's killer stuff.

I guess I'll just have to put together a how-to guide and sell it to the million dollar launchers out there

yeah, those cloud/grid set-ups are really cool... i believe they're stateless though, so one has to design around that.

[toddperk]

I got GCD 2.0 and can't seem to login...My account got approved and I got the email sent to me but I'm still getting this message when I try to login.

"Your GCD account is being provisioned. All accounts are verified manually. Please stand-by for a confirmation email from us as soon as your account is ready for use."

Anyone else getting this error? I don't see a help desk or support page either to contact..

[nicheminer]

Todd

I've got the same issue as you.

I bought about 15hrs ago and still no access to my GCD membership account.

If you email to support@gcdetective.com it gets bounced back with the message, no such email address.

Chris said he had servers all over the world and had done extensive testing to make this launch work.

Unfortunately it seems to have done a big fat nosedive.

What marketers don't seem to realise is that the customer is king, not just a cash cow.

If these things happen - OK - **** happens. But it is vital that the customer is kept up to speed on what is being done to rectify the unacceptable situation.

After the great videos and emails that Chris put out - I just want access to my purchase.

Graham

[calliope]

The irony being that GCD itself runs on a cloud of ~15 servers to distribute the load.

Originally Posted by entrepenerd

I'm still completely amazed that I haven't heard anyone talking about doing a launch using Amazon's EC2 and CloudFront service.

[DigiTips]

Originally Posted by ProductCreator

Of course it was fake, this happens all the time.

I mean, if they were all real then the no.1 priority of any launch would be "check server will handle traffic on day 1 without falling over".

Nobody buys this BS anymore.

EXACTLY RIGHT! - Another trick to get more buyers...come on guys, if you had a product launch, you would MAKE SURE your host could handle the traffic you was expecting....what kind of fool wouldn't??? (Unless you wanted that to happen of course..)..

but then who are we to hiss...how many server crashes have i had....or wish i had because of a product launch...hmm....NONE...

[JayStarr]

Well I am glad to see I am not the only one who realises this fake.. incorrect link BS on the mail outs.. it pisses me off.. I no longer unsubscribe from these morons I simply hit the SPAM button.. enough people red flag these mailers messages as SPAM with Gmail etc and pretty soon their delivery rates drop off.

As for the server crash on GCD I doubt its intentional, I mean look a number of people will wait a little time then its refund central.. plus all thos people who really get pissed and get their c/card to do a chargeback..ohh.. now you are losing your initial payment + a bank fee + a merchant fee + good luck trying to convince your merchant a/c provider not to terminate your account.

All in all I think if Chris could simply post a simple message on his home page saying.. "hey we are trying to get access restored sorry but please be patient it will be worth it etc.." and then simply update it every 12 hrs or so.

As it is hes dissappeared and the only communications set up is his automated messages preset to mail out.. already trying to sell filsaimes new product. It stinks.

Maybe Chris is in Mexico for a reason.. doing a runner.. maybe not.. but def poor CS.