Hello
I vowed I would never post something like this in a public forum and thos eof you who know me know how unlike me this is, but I have had another run in with Rackshack support and my head is about to explode.
Here is the e-mail I just sent to the VP of Support.

All I can say is I am glad I am not using them for our main brands.

Gordon




Hello
I wish to make a serious complaint about your quality of support
We have a server not responding to HTTP for the past 5 hours.
I entered a trouble ticket and got this sent back with no explanation:

from server history logs:

968 pico hosts.allow
969 ls
970 cd virtualhosting
971 ls
972 cd mappings
973 ls
974 pico sendmail.domainmap
975 ls
976 pico smtp_relay.domainmap
977 ls
978 pico domainmap
979 ls
980 pico sendmail.domainmap

------------------------


What has this got to do with the server not responding to http requests?
Those refer to me LOOKING at system files 20 hours ago to find out why mail
was not arriving
in a domain account.

The implication seems to be that I broke something.

Lets look at this logically:

1. If I had broken the server it would have stopped working 20 hours ago.
2. Checking those files was essential to trace the mail delivery issue (they
were not modified - just looked at).
3. Even if those files HAD been modified they would not have resulted in the
current problem.
4. As you do not provide managed servers it is ESSENTIAL that we look at
these files as part of the server managent process.

If you are impying that we broke something then SAY SO, don't send cryptic
messages and close the ticket.
You might also want to suggest a remedy.
You might also want to provide a reason why looking at these files would
cause this problem.

You have used the bash history log to close the ticket not solve the
problem.

This is completely sub-standard service and an incident which seems likely
to affect our continued business relationship.

:D

How about the network.
Can you connect your server via telnet or SSH?

Just broken for Http?

:D

Gordon,

This may be clever of them, but it is by no means customer friendly.

And if every service provider looks at the fineprint before solving a client's problem [or even uses it as an excuse to avoid support] it would be a sad day for customers indeed...

I am of the opinion you should have been given the proper service and ensured that your problem is solved

Balaji

I can connect via SSH and connect out from the box to other servers on other networks so its not a hardware fault.
I had a similar fault before and the guy who fixed it said it just needed an httpd restart.
Obviously I have tried that and a reboot.

Problem is they don't even offer paid support.
Most places will let you have a programmer for $50 - $70 per hour.

What annoys me is that they say its not managed hosting but when you manage the server yourself they claim that voids any support.

Not good.

Gordon

Originally posted by GordonH
Problem is they don't even offer paid support.
Most places will let you have a programmer for $50 - $70 per hour.
One of the main reasons i'm cancelling the server i have at rackshack.net soon

I think everyone now knows that support isn't a word in RackShacks vocabulary!

My server went down for over 48hrs..........why? I simply asked for more IP addresses! All support tickets were closed without reply or ignored. I lost a lot of money and never got an explanation or apology.

I learnt the hard way to avoid RS no matter how enticing their prices may be!

Former customer here...this is not a RS bash, it's just the way I see things...

They offer a price that can't be beat, but you do get what you pay for.

I also love the fact that they lowered their set fees to $99 for a couple of days, I wonder how many existing customers (like the guy that signed up last week) they pissed off during the process. That's one of the reason we moved away from them, running these offers and changing their policies about customers trying to upgrade their equipment.

I'm sorry, adding 800 servers in a single month, something's going to break on the back end. I seriously doubt that kind of growth was planned, at least not from a support perspective.

Finally got it fixed:

------------------------
your error_log for apache was too big..(over 2 gigs) this was causing
apache to fail.

note: you should use less or more to view files....not pico or vi..thats
dangerous.

--webtech xxxxxxxxxx
------------------------


Why did they not just say that in the first place insterad of trying to fob me off.
Thats what really annoys me not the fault - faults happen.
I never suspected that fault as I have cron set to delete log files occasionally top preven this.

At least it is now working.

Gordon

How nice of them to be condencending in their final response. That company really needs an attitude adjustment, they act like they are the only providers out there offering their breed of service, like they have a monopoly or something. Looks like up and coming competition will either bring them to their knees or show them that they need to shape up.

BTW, don't use your blow dryer in the shower, it's dangerous. Sheesh.

Originally posted by DizixCom
How nice of them to be condencending in their final response. That company really needs an attitude adjustment, they act like they are the only providers out there offering their breed of service, like they have a monopoly or something. Looks like up and coming competition will either bring them to their knees or show them that they need to shape up.

BTW, don't use your blow dryer in the shower, it's dangerous. Sheesh.

Well they are the only one offering 400 GB and this and that for $99 a month :)

I find it funny people complain about the $99 a month and they get no support. They are unmanaged boxes with 400 gig of transfer. If you want your hand held goto rackspace or someone else.

I would be more concerned with how they respond to their customers than their actual support.

Their cryptic response was rediculous to begin with. Why didn't they just say they couldn't support it instead of cutting and pasting irrelevent log entries? That's what I find humorous.

And then they respond with a remark saying to use less or more instead of pico or vi to view files because it is dangerous.

No support, fine, but to be rude about it and condescending? No thanks.

I'd have to agree, if you dont want to manage servers, or support more then hardware/network just ask "can you connect to the server?" 'yes' "good, network is fine, thats all we can do for you sir, im sorry". Thats still a slightly a**hole response by most measures, but is crystal clear.

Not to be rude, but it is clear that our techs, even though it was clearly an error unrelated to network or hardware, reviewed and resolved the customers issue.

Eventhough our support is quite limited in scope, we frequently repair software/configuration issues that are the result of end user error.

I fail to see how the response above encouraging use of a particular command was condesending. It was clear, consise, and straight to the point.

Just a few minutes ago, we had a guy screaming in our chat room about a server down. He had been working in the server, edited a few lines, made a type, and blamed us for it.

It was not our fault, but we repaired it anyway at no additional cost.

We have our fair share of issues, that is certainly true. What is not fair is to issue a blanket statement "Rackshack support sucks". It is unfair and untrue. We sell dedicated servers, not managed servers.

For customer who cannot manage their own server, we might not be the best place. If on the other had, you generally know your way around Linux and just need a little help here and there, we come through with shining colors.

Robert Marsh
Head Surfer Rackshack.net

Originally posted by GordonH
Finally got it fixed:

------------------------
your error_log for apache was too big..(over 2 gigs) this was causing
apache to fail.

note: you should use less or more to view files....not pico or vi..thats
dangerous.

--webtech xxxxxxxxxx
------------------------


Why did they not just say that in the first place insterad of trying to fob me off.
Thats what really annoys me not the fault - faults happen.
I never suspected that fault as I have cron set to delete log files occasionally top preven this.

At least it is now working.

Gordon actually that 2gb file limit problem is the first thing i check these days.. you'd be surprised at how many servers i looked into that ended up with this exact problem!

Robert,

No offense intended towards you, it's just that you sum it up yourself:If on the other had, you generally know your way around Linux and just need a little help here and there, we come through with shining colors.Your techs automatically assumed that this person didn't know what he was doing, and instead of explaining why they couldn't help they just pasted an irrelevent portion of .bash_history. The last message comes across to me as condescending when your tech scolds this person for using pico instead of less/more to view the files. If you are expecting your customers to be experienced sysadmins and to have general knowledge, then give them the benefit of the doubt and assume they know what is best for them. Remember, it took a message to some vp to get any kind of action at all, that (to me) is not "shining colors."

With all due respect, that was certainly no scolding. Simply a pointer for the customer. That's all. Nothing more. Nothing less.

You seem to be reading into that statement something which clearly is not there.

In the end, this customer was assisted with a problem outside the scope of our support and even given a pointed that could help in the future.

Robert Marsh
Head Surfer Rackshack.net

thats dangerous.You're right, maybe I read to much into that, but I get a vision of a school principle waving their finger in the air "Young man that's dangerous!" or "You'll shoot your eye out!" If you say it wasn't intended to be a scolding, then I'll take your word for it and concede. The bottom line, as you say, is that it was resolved in the end.

Robert,

Is telling a customer that the problem is the size of the error_log being too large that far out of your scope of support? I mean it's not like that answer requires a whole lot of debugging or something.

I just thought I would add my most recent experience with RS. You can read about my last experience by doing a search on the forum for rackshack. Now, after purchasing a Dial up internet connection for the sole purpose of purchasing a dedicated server at RS I managed to get to the sign up forum without being banned from the server (luck was on my side, it seemed). I ordered one of the $99 setup Intel P3's that are supposed to be equiped with 512 MB of ram. I quickly found out mine only had 128 MB. After pointing this out to RS in a support ticket, they were quick (well a couple of hours quick) to point out what I had already told them. They updated my ticket to suggest the ram in my server was inadequate. Nothing else was added to the ticket at that time. No, "we will get right on it", or "we will have to wait until 9 am for customer service to approve us adding more ram". Which was the case, I found out through pestering the tech's in the IRC chat room. Why could they not tell me this in the ticket? I then got a responce to the ticket this morning saying: "Server had 512MB ram in it. I swapped it with 2 sticks of 256MB. Total
of 512MB. closing ticket." Now, my server says 496732 KB.. which according to me is about 485 MB. Perhaps my calculations are off, I don't know. All I do know is my host name has been changed, and I don't know what else has been changed, but its time for further investigation.

All servers (excluding Cobalts) come with 512. Your server had a 512 stick in it when you bought it. Apparently, the Ram was bad so we swapped out the Ram.

Customer service is not involved in any hardware replacement issues where there is a known problem with the hardware.

Your problem was quickly escalated to a tech that could replace the Ram, Ram was allocated, and the Ram was replaces with 2 256 Mb sticks.

We will occasionally have hardware issues but yours was fairly quickly resolved especially considering the volume of activity yesterday and today.

We try to treat every hardware in a manner that we would like to be treated if we had a like problem.
What you don't see in the notes is the escallation between departments (such as between Tech and Data Center).

As a courtesy to our customers, we automatically send an email everytime the ticket is viewed, escalated, or closed. The notes in the first ticket are a techs acknowledgement to our staff of the issue. They are not meant to be a communication to you. The email sent at this point is meant just to let you know that we did look at your ticket.

The next communication point was that we swapped a 512 stick for 2 - 256 Meg chips. At this point, the problem was rectified and the ticket closed. What more can we say than a 512 stick was replaced by 2 - 256 Meg chips?

The information was presented in a straightforward, simply manner in a fashion that allow the tech to move on to the next ticket.

So, as intepretation, the customer opened a ticket that said he had a Ram problem. A web-tech looked into his server and acknowledged a problem with faulty ram, the ticket was escalated to Data Center and the Ram was swapped out, noted in account, and ticket closed.

Teck-
Our committed scope of support is to ensure that the network is operational and that the server is able to access the network. In reality, as was the case with this customer, we generally step outside those support boundries and help our customers in almost any way possible with the exception of installing an dconfiguring software.

Robert Marsh
Head Surfer Rackshack.net

Originally posted by teck
Is telling a customer that the problem is the size of the error_log being too large that far out of your scope of support?
Well, this doesn't exactly sound like something that would be included with an unmanaged server. I see now why dialtone (that is right isn't it?) makes you agree to being charged for support when you submit a ticket. While this makes people uneasy, I understand why they do it.

I realize there are a few things being discussed. One being that RS provides unmanaged servers. Two, that people wwould like paid support options. Three, Haze and others posting faults in true support requests to real problems (real problems meaning problems with the hardware/software either upon install or in general).

Maybe the answer to some of the problems is to find a few admins willing to work on a per incident basis to support these servers? I suppose RS could do this, however I was thinking mainly about people who lease the servers looking for these people.

It seems dialtone's 'agree to pay' system actually is the best way to go, however I'm not sure RS wants to provide this type of service. I also wonder about the 'lease RS through me' type deals that a few members have posted (sabaro....errrr something, Dennis? I think, and another one, sorry the names aren't coming to me at the moment). 'Lease through me' at a slightly higher rate might give the best of both worlds?

I'm really just thinking aloud here, sorry it probably doesn't make much sense. I'll stop rambling now...

I understand where you're coming from. Probably instead of having RS offer more bandwidth, cheaper prices, try to focus their attention on servers that are a bit "more managed." Maybe lower their bandwidth allotment by half, raising their prices a bit.. and provide some sort of deal for people who need that extra help. The "cheaper" servers can still exist for those who would just want the totally un-managed server.

Teck-

There are lots of great business models out there. Ours is pretty siumple and the price point coupled with agressive advertising delivers a CPS (Cost Per Sale) that makes our model work.

Other companies that are currently agressively advertising, such as ** fill in blank of print ads for hosting that you have seen recently**, probably have higher CPS. We subscribe to a clipping service that monitors print publications as well as a tracking service that monitors spending by companies (advertisers) that they track. Couple these with published run rates on new servers and paterns start to develop.

That pattern suggests that "subscriber acquisition" is a cost center, not a profit center. As a "cost center", those costs have to be recooped from the MRC, resulting in higher monthly fees to start with even before you add costs of premium support.

With a lower price point on our product, it reduces the barrier to entry, resulting in volume selling thereby lowering acquisition costs and turning subscriber acquisition into a profitable activity.

I have delibertly avoided having a part free/part pay support policy as I don;t want to have an argument with a user over what we charged them for a particular support call.

Back on your suggestion, and I am sure that it is a good model. Dialtone includes less bandwidith, prices that are double (at least), and seems to do ok. If their current run rates are what they last published on their website, I am sure they're doing well.

Our model is different. I might even compare Rackshack to Walmart (high volume comodities at low prices) as opposed to, say, Herrods. Herrods would never want to be considered a commodity.

We've taken dedicated hosting down to lowest common denominators: space, power, server, and bandwidth. By building systems that create efficiency such as sales strictly online, automated provisioning, limited SKus, and buying in volume (bandwidth), we have a model that works for us.

Other will have models that work for them. I'd like to server the masses, not the Fortune 1000.

Even further up the scale, there are models that are concieved on a very FEW large managed services customers. Navisite is an example but there are others. All you have to do to know their models are flawed is to look at the 10k filings.

I guess what I am trying to say is that every company will have its own take on this business. We have our own, and for whatever anyone thinks, we've developed as a volume leader in dedicated hosting market in just a year with currently over 4000 dedicated servers, over 7000 virtual accounts, and now over 320 employees. Heck, we've done over 850 dedicated servers this month so far plus a few hundred virtuals and about 12,000 new dialup accounts.

Dialtone and Rackspace are great companies too but they serve a different market segment, more middle ground.

I seem to get emails every day with a suggestion that we should do this or that. Build this server or that server. Offer this control panel or that one on this server of that one. Offer a support package with this support and this commitment.

Every time I even think about adding a new this or that, I try to remember what the "model" is and how it works: Volume.

We will always have distractors including those that love to hate us for whatever reason.

In the end, while we will continue to "tweak" our model, we'll most likely stay close to our roots : comoditized dedicated servers at market setting prices.

Open forums, such as WHT an dour own forum, tend to bring out more "negative" than positive. It's just the very nature of the beast. With our volume compared to others, it is just natural that there will be more folks airing "dirty laundry".

Robert Marsh
Head Surfer Rackshack.net

For what it's worth, I derive a more positive image of Rackshack than negative simply because of the volume of customers you have and the limited number of qualified complaints. The only thing that really scares me away is the lack of physical access to the server and the fact that it's inevitable that some day I will need that.

Maybe I'm just paranoid. Quite possibly so.

I think of RS as a colo..... except I don't have to ship them a server!

I admire RS crossing the support boundaries as Robert says they routinely do. But... I think if you can ssh into your box then it's the customers issue. I'd look for a linux guru as Chicken suggests before calling my "colo's" support desk. I used to think that RS should be responsible for:
1- network efficiency
2- h/w repair/replacement
3- hard reboots -*could be elimnated w/ remote insight boards :)

I guess I have to add one more... handing out IPs... so thats 4 things. I guess I'm just a jerk about these things.

As long as the customer can login their dedicated server, and the OS is running fine, rackshack business is done. It's up to the customer itself to administer the server.

To ID a problem is a time consuming process, Rackshack did it for free, that's great service!

"I guess I have to add one more... handing out IPs... so thats 4 things. I guess I'm just a jerk about these things."

OK. So this is one of those things we could do better and faster. We are working to implement an automated system to speed things along.

These things develop along the way. There does not seem to be one magical piece of software to make everything on the backend work just right.

Robert Marsh
Head Surfer Rackshack.net

In the end, this customer was assisted with a problem outside the scope of our support and even given a pointed that could help in the future.


Why does your trouble ticket form have all those problem categories then?

Gordon

We alwasy try to help in those areas when we can. It also allows us to cassify tickets so that they can be routed to the proper department.

Robert Marsh
Head Surfer Rackshack.net

Robert,

I'd like to get a box from your company mainly just to play around with it and learn etc but your order form has been down for ages now. Any ETA when that will be operational?

Parts of Australia IP space remain blocked due to fraud. Just send in the fax signup an dour folks will get you going.

Robert Marsh
Head Surfer Rackshack.net

I thought that might be the case. Gee, I knew parts of Asia was rough for credit card fraud but Australia?? Hey, we're the good guys :)

Sounds like they have blocked the whole of 202.
If you look at the thread under running a web hosting biz you will see a breakdown of 202. whioch avoids blocking Aus and NZ

Gordon

Gordon-

What an awesome thread with a LOT of great info. Thanks for pointing it out.

Robert Marsh
Head Surfer Rackshack.net

Originally posted by headsurfer

OK. So this is one of those things we could do better and faster. We are working to implement an automated system to speed things along.

Don't take the IPs as a complaint.... it's just one of the 4 services that I think an unmanaged provider (or a colo) such as RS has to provide. I think everthing else should be done for a fee.

Originally posted by teck
Robert,

Is telling a customer that the problem is the size of the error_log being too large that far out of your scope of support? I mean it's not like that answer requires a whole lot of debugging or something.

I know this is an old message, but I felt the need to reply.

Read Rackshack's TOS (http://www.rackshack.net/aboutus/tos.asp), Part 11: Support Boundaries. (This section also applies to another thread that is locked.) Rackshack's resposibility lies only in the hardware itself. The fact that they are willing to help out with software issues is a good thing, but it is not part of the service that each and every person who leases a server agrees to.

Sorry
I don't agree with that.
They are supplying a server with Ensim Web Appliance and they must take responsibility for maintaining that software.

Even Microsoft provide support for Frontpage after all.

There are two issues here:

1. Not providing tech support.
2. Being rude to customers.

Point 1 I can cope with.
Point 2 makes me very angry.

Sending that message back to me was rude and unhelpful.
They didn't even say why they were sending a copy of the bash log.

Also, its actually the second time they have pulled this stunt.
Previously they had claimed I had edited the httpd.conf file on another server because they had seen it opened two weeks previously in Pico.
No explanation of how that would relate to an unrelated fault.

If I had broken the httpd.conf I would have fixed it.
That again was an Ensim problem and they used a previous entry in bash history to avoid the issue and close the ticket.

That looks like a policy, and not just one persons off moment.

Gordon

So, are you saying that even though you agreed to the AUP/TOS, that it should not apply to you anyway?

Anytume there is a contractual relationship, especially where you must scroll through the document and click for acceptance, you have agreed to the terms.

If you did not agree to the terms, you should not have proceeded after reading them.

Once you read them and clicked to agree, you were bound to them.

The response was NOT rude. It was straight forward and to the point. Heck, even after we fixed what was a user problem, we even gave you a pointer to help you in the future.

We stepped outside of our agreement and went the extra mile. I just cannot believe we're getting hammered for doing MORE than our contract (AUP/TOS) require.

I've said it before and I"ll say it again. We'll do almost anything if we are able to help our customers just as we did in your case.

Robert Marsh
Head Surfer Rackshack.net

Robert
Please re-read my post (directly above yours)

I was NOT complaining about the tech support issue.
I was complaining about your companies use of Bash history log on two occasions to close a support ticket without explaining what the relevance of that log entry was.

If you look through my history on this board I think you will find this is the first time I have ever criticised any supplier in public (heck, I lived throught the VDI Friday and never voiced a complaint in public as the people involved were totally upfront and were doing their best to rectify things).

What annoyed me was the sending back to me of a list of file viewings which bore no relation to the fault or even took place in the same time frame.

The first time I shrugged it off but the second time I felt insulted and I have never felt like that about any supplier - ever.

Here are some comments I hope you find constructive:

I recommend that you change your support ticket to make it clearer on a daily basis what your support includes.

You also need to make support available on a paid basis
(surely at the standard $50 - $70 per hour you could make a profit?) We only use support like that occasionally but its essential when you are dealing with unfamiliar software.




The real problem is Ensim. Its complex and if something goes wrong its very difficult to put right because of the way Ensim structures the file system.
Its not just an add on, it really dictates a lot of the system operations.

We may continue to use rackshack for some purposes but we will need to install our own software so we know where we are (something which was not possible a few months ago).


Gordon

First - One of the problems on the server that I cancelled with RacKShack (Ensim and I will bet GordonH has) was error_log file crashing the server due to the Bandwidth Manager writing lines several times a second to it. Assuming his problem is the same.

Since RackShack charges to setup a Server with a Control Panel that they selected (and hopefully trained on, tested, etc..) they should support the problems with it.

RackShack sold these Severs (1000's of them) with Ensim on them for several months knowing hostname, bandwidth manager and other significant problems exist. In each instance when I requested assistance in correcting them the techs supplied no information. I found answers on the Ensim board.

Second - I have had many experiences with RackShack Techs being rude and closing tickets assuming I knew nothing.

I have mixed feelings on RackShack support. Up until 2 weeks ago, I would have praised them to the stars. But I bought one of the Ensim boxes from them during their special 11 days ago, and I still can't upload a site onto it.

There was a problem with the RAM in the box, and they did a switch. But there was also apparently a problem with the file configurations being so off that I had two tickets closed because they said I had changed things around. And each time, I had to call in tech support to verify that the changes were so extreme that they couldn't even be done remotely.

After the RAM change, I still had problems, so I requested a Restore (which a phone tech also recommended). Hey, if things are that messed up, I don't want problems patched individually, particularly when I am going to start putting multiple sites on the box. My request was ignored, and they tried to fix the FTP, which they didn't. I emailed Mr. Marsh requesting a Restore last Friday. I don't know if he responded because, get this, my main email which is located on my other box, my RAQ, stopped working Sunday, and still isn't fixed. That was part of a problem with my root being filled up, which turned out to have something to do with a partition problem with the box. But apparently it is something more.

My experience with Rackshack is that the support personnel are usually very friendly & helpful, but the people who actually fulfill the trouble tickets take less of an interest in seeing the problem resolved than they should. They are very quick to say that you have changed config files and thus voided support, but at least if you press the issue that you haven't, they do try to resolve the problem for you.

My position is this: I am not that technically literate, and I only use my Rackshack box for putting up static pages. If I was building dynamic sites, I would probably look for another hosting company with more sophisticated offerings. I don't touch ANYTHING on the server. I add and update sites, and that is all. Under those circumstances, if something goes wrong with the server, and that is actually very rare, I really want help on the problem.

Rackshack is offering a nice deal, and I hope they make a ton of money off of their business. But I do see a problem when I go into their chatroom and people are asking about advanced configuration issues or discussing things like PHP Nuke. I'm glad people are getting technically creative, but I think it leads to problems re support that I don't think Rackshack has properly addressed.

If Rackshack wants to be the Walmart of dedicated hosting, then they are going to have be a little bit more responsive to people like myself who just want a bare box--which I think is their primary market--and evolve some different options for the individuals who are skilled in Linux and will be installing MySQL and everything else. I know I could care less if my boxes even had PHP or Chilisoft.

There is an obvious solution to make their lives easier.
Ditch Ensim and do Cpanel.
Regardless of what people say about cpanel it does update all the software every 24 hours so there is no need to keep patching the servers manually.

Gordon

The problem with Cpanel is that its high price tends to exlude it as an option for the first-time-server owner. If I only have $99/month to spend on a RackShack server, I probably don't have an extra $50+/month to spend for a control panel.

Rackshack would only pay $25 per month for cpanel (list price for NOC's).

Thats not much more than they are probably paying for Ensim.

I am sure they could force the price down if they were buying even more licenses.

Gordon

Now that would be interesting if RackShack could put togehter a deal to offer Cpanel at an affordable price... I wonder how much could be saved with a deal to buy and offer 10,000 Cpanel servers :D