Major Network Outage.. (resolved)
Posted (March 28th, 2007 at 5:33 pm PST) by JoshWe just had a drop in our network for about half our servers.. if you can’t access your site or email now, please hold on.. we are looking into it and will update this page as soon as we have more details..
7:15PM PST
We’re still actively trying to diagnose and fix the problem.. it seems to possibly be a physical problem with one of the fiber cables between two of our data centers. We’ll post further updates here as we progress. Thank you very much for your patience and we apologize for the downtime this is causing.
7:20PM PST
Scratch that.. it turned out to be a failure in one of our 10 gigabit network interfaces. We’ve swapped it out for a spare one and thing seem to be working now.. there may be some residual problems where some servers need to be rebooted, however the root cause of the problem is fixed. Please let us know if you are still experiencing any downtime.
Our apologies again..
135 Responses to “Major Network Outage.. (resolved)”
Boba server seems to be down and out.
couldn’t ftp, can’t bring up website.
something [major] is going on
ARGHHHHH. Is the entire DNS system down??
I have accounts on 2 servers and both are completely unreachable.
How many times is this going to happen. WTF is wrong with your network!?!?!
well there isn’t a day when dreamhost is not down! we need at least 50% uptime. please you should be able to at least keep the sites up for half the time!
I already moved all my mail elsewhere and will probably move the web also. I will keep my account bc the amount of space and bandwith makes it a good place for off-site backups
all mail and mailing list programs aren’t working on my account.
Dreamhost has been really shaky this year. 
And if you have “major clients” coming to your site, move to vps or dedicated..
wow, even dreamhoststatus.com was down (it only showed a one-line message about the mail server)
looks like things are back to normal, thanks DH!
Thanks for fixing this quickly, Dreamhost. A little downtime now and then doesn’t affect the fact that you’re a great company!
Who spilled beer on the server *this* time? 
I feel for these guys — we really depend way too much on them. I honestly think that they should lower their offers, and report true SLA.
Thank god we’ve found a Dedicated Host. 200$ a month? Sign me up.
@matt: I think they changed it to a one-liner just to avoid a gazillion comments in the threads. In a situation like this, I probobly would have, too!
“little downtime”? ahgahahah
I’m glad you fixed it quickly and I’m a big supporter, but this is getting a little ridiculous. How often do you guys do, like, preventative maintenance?
I’m just gonna say that this was pretty impressive. The initial problem was reported on dreamhoststatus about an hour ago. 30 minutes later we have another update and 30 more minutes and the problems resolved. Shit happens, and they cleaned it up fairly quickly.
Oh and I second Josh’s comment about “major clients”. If you’re building a business you should shell out for more than shared.
DH, Your services are awesome, but whats with the constant network issues guys? Will these calm down soon?
They can only hope.
Don’t get me wrong… I love DH. But Shared Hosting is paid for in volume. I was ready to shell out 400$ for a DH Dedicated Host, you guys are a _SUPER_ good deal.
Just that dependability comes at a price, and I’m glad you guys are offering services. But your sales material should be updated with real SLA.
One final thing… DH Dedicated Hosts are on a wait list > a month. Hope to come back to DH.
Ok, I understand, shit happens, but not ONLY shit happens! I started 24/03 and until now there is no 1 day without problems! I use many other free hosting services and I never see anything like this happenning on there… I expected at least this of DH!!!
it’s 23 bucks for the first year and 8 bux a year after that .. you can’t get too angry :]
For people who say they have lost, business, or lots of money, if you lost that much money or that many customers you should invent in dedicated hosting, and not shared hosting. It only makes sense, if your site needs to generate $$$ then you need to make sure its always on. But really if you business is that good, your customers will wait… Trust me.
Be happy Dreamhost communicates with us when they have problems other company’s will not post or admit troubles in a network most hosts don’t, and make you pay for a 30 minute long distance call to support to find out about a system error.
Crazed when you have as many servers as they do its kinda hard to keep up with everything.
First major outtage since ive been with DH and i knew they would have it fixed in no time.
@mike: nice… Now I understand the DH philosophy… Its cheap, don’t bother… Ok, I think thats not a good way to go.. Coke is cheap and I never get disapointted…
@Eu, but coke makes you fat thus causes you to pay for health club fees which are not cheap by any means,
In any case…. I signed up this afternoon and sure it sucked losing connectivity right in the middle of a move (I was stuck in vi server side when the poop hit the fan) but I managed to move 3gb of data and get everything up and running in a matter of 3 hours and that includes the downtime.
@Jason: Perhaps. Realize that I’m as big a fan of Dreamhost as you can get, but if a monthly once-over of things can prevent problems like this, then everyone will be a lot happier.
Just because something doesn’t cost an arm and a leg doesn’t mean that we should expect sub-standard services. Whether I pay $10 a month or $50 a month, I have purchased what’s written on the sales page.
I too rely on my websites for income. And no, I shouldn’t have to fork out hundreds for my needs. All my sites require is dependable uptime! THAT is not too much to ask of any host, shared or not. I don’t host big videos, sucking bandwidth… I don’t go all crazy with huge file uploads and downloads… no load spikes. If I had those things, I would invest in something more robust.
Dreamhost makes promises in their sales materials, yet for a while now, there have been serious issues. Has anyone been offered any kind of compensation for the problems? A free month hosting? $50 on credit? Anything? Seems to me I’d do my best to keep customers happy while they figure out what the hell is going on with their hardware.
Oh - and my sites seem to have returned, but email is still down. Wonderful.
Dreamhost - appreciate the services you guys have, but could you please explain to all of us why uptime has been so difficult to maintain?
If we had an explanation, we could make a choice about whether to stay with you.
Without this, I’m getting pressure from my org to switch.
Please help me convince them things will get better!
@Yvo, I drink coke zero! 
Interestingly, with this issue our server was OK, contactable, ssh in and so on, but the database server was not. So since all software we use is DB bases, things didn’t work and got 500 errors.
@eu
Ok, coke zero causes cancer and that isn’t cheap either!
OK - Dreamhost just called back - appreciate that, BUT: the tech support person told me that for the next month they will be moving servers around and that consequently we could continue to have problems.
PLEASE: if this is the case, please inform us ahead of time so we don’t depend on your services during this time.
Jesus, you’re retarded. Be more reliable.
Dreamhost should drop the “strictly business” plan. My personal cheaper server is up way more than the one used for clients. *sigh*
My boss is pissed!
Why should dreamhost make you money if you aren’t making them money? If you are depending on 100% uptime then you can pay for it, otherwise, you can have fun with some downtime when it happens and not complain about losing ‘clients’. Though, really, I don’t know what kind of a retarded client can’t view a website at ANY time of the day and only during a specific hour. Also, why do you not have a local copy if you are, for some reason, showing your client your website in person? Don’t you think it would wow them more if your site was loading with zero lag?
These are some things that people who don’t understand what a massive amount of service they are getting for the low price can think about before they post on here about their made up troubles.
Didn’t even notice the down time and I was checking my site on and off throughout the day.
Thank you DreamHost for dealing with this quickly and efficiently.
Any clues as to why the interface failed though?
Ah!
All of my sites working with fast loading performance.
Thank You Dream Host Team. 
I have two sites running. One is on dreamhost and the other is on globat. The yearly payment for shared hosting is comparable between these hosts.
I was pleased with both dreamhost and globat till few months back. Now both suck.
1. Dreamhost - Too many “so called” network issues.
2. Globat - The network is super. No downtime/issues so far. But they charged me $40 for the so called spam shield!!
Thanks for the updates. Your shared hosting is still the best. I have another blog on dreahost and whenever it goes down myphone rings off the hook. I have seen more frequent outages here than when I hosted my web server in my apartment, but overall I’m very pleased with your services (but why no scp or secure ftp?). Cheers
our site is off again, was back for a while there.
Hey, don’t think us dedicated people get better service - in fact we get worse. I wish we never got a dedicated DH server, let alone two. The panel only slightly works with the DH Dedicated solution, and downtime is much higher on those servers than our shared accounts.
@John:
Dreamhost doesn’t make me money, I make my own money.
AND I also make them money - by paying for hosting services and with bringing a multitude of clients along for the joyous ride. (currently almost 300 customers)
I don’t expect 100% uptime, but certainly FAR better than the recent 40% or 50% I’ve been getting.
And as for “retarded clients”, when you deal with REAL business people, they have schedules. You make appointments and drive sometimes hours to meet with them. You show them their sites WORKING, online - all connected to their databases and email and blogs.
And like I said - the price I pay is irrelevant. I purchased a package that included far greater uptime than what we have been getting for quite a while.
Do you really believe things would miraculously be all better if we were charged double what we pay now? I HAVE paid for dedicated service - it’s absolutely no better.
This outage took out half of the servers.
By now, I’m used to seeing the same old exchange of comments:
“I’m losing money!”
“If that’s the case, why not go dedicated?”
However, in this instance, it appears even people who have paid more money for what really ought to be a better service are suffering.
I’ve also noticed that the average server load on the server I’m on (Formosa) is constantly indicating that the CPU is over-burdened.
I think all of us have a reason to be upset over the latest outage. On the other hand, you can’t predict in advance that a network interface will up and fail. Things happen and it was certainly inopportune for me, because I was writing a long update for one of my sites and I had to stop and wait.
I can see where a commerce site would lose money when these things happen, but how do you avoid system breakdowns, involving equipment that isn’t built by your host?
Peace,
Gene
My sites are all down. Again. I’m starting to look for an alternative provider. This has been too much this last month.
It is cheap, but it doesn’t deliver. You pay peanuts…
My sites are still down, why no update on this?
My sites are still very slow.Sometimes,all sites would be down.
I’m losing money now.
Uhhh…bad day for this, very bad. Some updated print files never reached their destination and now it is too late - a large job is going to print wrong. Time to start shopping around for a new host.
Please, people who’re shouting “if your site’s important, don’t use shared hosting, use dedicated hosting” - stop being assinine. Dedicated hosts at dreamhost are affected by network and power issues at well - the infrastructure is just not great.
As for “what do you expect for $7/month” - well, I expect a better infrastructure than my home DSL modem and residential power; that’s why I shell out ANY money at all, rather than running it off a home server. It’s the whole idea of shared hosting; pool your resources so you can afford decent infrastructure (and some sysadminning).
I haven’t checked if my sites were affected yet (no clients yelling at me, but they only use e-mail anyway) but these apologist arguments are getting really old.
My site is off, hurry up please. Is in the morning in Europe and I lose visit for my web.
@gene
“I can see where a commerce site would lose money when these things happen, but how do you avoid system breakdowns, involving equipment that isn’t built by your host?”
Redundancy and failover. If one GigE port breaks down at, say, a AOL/Google/Yahoo datacenter, it doesn’t take out half their service. Same for powerstrips, power outages, network uplinks, etc. Dreamhost claim to be unable to fix the building’s non-redundant power infrastructure, but apparently there’s not an abundance of redundant power supplies and network connections within Dreamhost’s own domain either.
They should really hire some one experienced to come in and do the math.
Seems that my wordpress instillations have snappy response times now - the admin pages would take upto a minute to load before. Did this fix something that was wrong the major downtime?
Perhaps the fact half their servers went down meant less traffic to the sites, thus making your own more responsive?
I’ve been with dreamhost for about a decade. I’m a very happy camper making some serious money and all my sites are hosted with you kids (except CrazyEgg & CellForCash & totaltrak obviously)
Man, I sure hope all these whinging asshats piss off to other hosting providers…
resolved?
why i still can’t connect to my ftp ?
much tooooooooooooooooooooooooooo slow with all of my sites, they are loading into the next century..
wow…i agree bad timeing….still my site wasn’t down long…. i thought something was really wrong because even the dreamhost status was down so i couldn’t check to see if anything happened. I am glad Dreamhost got it back up…i really hope they get there act together for the people who “rely on there website” to make their money. Lucky mine is just personal.
woof. stoopid end users.
looks like problen NOT RESOLVED!
i still can’t connect to my http://ftp. Domen - transparenthouse.com
@Vince: “How often do you guys do, like, preventative maintenance?”
Please don’t encourage them to tinker. Otherwise they’ll do something stupid like paste the wrong IP in the wrong field and create an infinite loop (yes, this has happened - and recently).
BTW, DH Status seems to stay up through thick and thin — apparently it’s not self-hosted? So where is it hosted? Whoever does it seems far more reliable.
Why would servers need to be rebooted if there was an upstream network issue? o_^
One of my old hosts had an “emergency message” page (where you could report severe outages.)
Annoyingly, whenever my site went down, so did that page.
@Roxanne: scp is available on all boxes.
@List Preston: hit the nail on the head!
@mike: how are you getting this super service for 8 bucks per year?
Thanks dreamhost. I just lost my job for choosing you.
@fivecentnickel.com BTW, DH Status seems to stay up through thick and thin — apparently it’s not self-hosted? So where is it hosted? Whoever does it seems far more reliable.
Dreamhoststatus appears to be hosted at http://www.fibernoc.net/fic an establishment that prides itself on having redundant uplinks etc. - but, they don’t do shared hosting, they sell fiber connections.
Can’t reach either of my DH hosted sites, hlcomic.com or notmydesk.com.
Thanks for the quick fix!
Some of you really REALLY complain too much.
Dreamhost’s lack of network redundancy is a little surprising.
Lisa Preston: I agree completely.
Please keep in mind that dreamhost offers different packages (not really a whole lot of a difference in the packages) so not everyone is paying $7/month. Many DH “business” clients are paying 10 times what other clients are paying and getting the same, if not worse, downtime.
OMG!!!!! It’s the end of the world, WE’RE ALL GONNA DIE…
…and it’s all gonna be DH fault!!!!
Let’s sue them for all the profits lost, all the wars and the planet getting hotter. If you’re gonna host that many sites on that many servers, you’d better have some COOLING a**holes! You’re warming up the planet you stupid retards! KYOTO anyone ?????
People who work at dreamhost are all rapists, I can’t believe I pay a Gazillion bucks for this crappy hosting hosted by sexual perverts. I’m leaving NOW! This has to stop…
nah,…I love those pages and network issues, That’s why I joined DH in the first place
Here’s a little reality check: First of all, understand that Linux servers used by Web hosts are basically gussied-up PCs with fancier and redundant power supplies, larger cooling fans, and the fastest processors from AMD and Intel. The hard drives may be more resilient than the $129 versions you pick up at Best Buy, but they still fail. They have to run 24/7, along with a complicated array of networking hardware, cabling and so forth and so on.
Equipment breaks down on a fairly regular basis. Sometimes it happens all at once, in which case lots and lots of people are impacted. Sometimes it happens in bits and pieces, which means only a few, those on a specific server, or in a specific node, are hurt.
I’d like to see greater reliability. On the other hand, nobody here has ever personally visited DreamHost’s facilities to see just where things might be improved.
More to the point, if you think other hosts don’t have failures, do a little online checking first. You’ll see they all do. Some are better, some are worse. But the grass isn’t always greener.
Peace,
Gene
As a sysadmin I must say that I run a bunch of servers in a Datacenter at my institution. Since we use a wide variety of servers we’re able to replace a downed webserver very quickly with no data loss. We control the switchs and we simply tell the switch that another machine is server that IP address. Then that machine’s webserver is told to serve the various sites. They all mount the same remote filesystem so they all have access. I’ve noticed dreamhost does have a fileserver so with a little setup they’d be able to tell their switches (if they bought real switches) which webserver is serving that CPU and with a tiny bit of configuring update that webserver’s configuration. The downtime would not exceed minutes.
This isn’t hard.
@sysadmin: Got roughly 1500 servers across at least two datacenters to manage?
Peace,
Gene
Gene, what part of redundancy don’t you understand?
Gene, what part of redundancy don’t you understand?
And do you have a full understanding of DreamHost’s network architecture and such? If you want to criticize, fine and dandy, and I’m not denigrating your experience. But you don’t have a lot of info on what they have.
Peace,
Gene
Weird, my site wasn’t down at all… a little slow, but still up.
Guess I’ve just been having great luck when most others have been down.
Well.. Redundancy is one thing and u can use that to prevent a lot of faults, but even though there is redundancy, it doesnt always work. Now.. what if the Interface before the Switch goes down and the switch array isnt contactable? Huh.. Now what to do… yes… redundancy is good and most certainly a must when doing a large array of hosting. But… the fault which it looked to be, one of the fiberlinks that interconnect the SPACED out and REDUNDANT hosting facilities, as I understand it. Still things can go down…. and to take another server and tell the switch to port over, yes that is doable, but is it doable for 100 servers at once ? Do you keep 100 Machines / VM Slots on standby ? And would that be Cost Effective in the world that changes more or less hardware wise from one moment to the next ?
I work for an large Telco that has large lines of fibermanagement and many many many Broadband lines out to customers, and I do fault handling management / Support on the “Frontline” / “SecondLine” tier and we have customers in Different SLA lvls… but sometimes there is a general fault where the situation is affecting 5 - 10k+ customers, where some off them has 24/7 Fault Support, but… when its the case of a general fault and the Fault Correction is started within the parameters of the SLA, then they are working on it. Yes.. if its an longer outages their is a fine to pay, exempt of course in Force majure.
I Did notice that, as soon as DH discovered the fault or problems, they where fault tracking / trying to pinpoint the problem, and also notice is, that they did get the problem fixed in an timely manner in correcting a problem on their Backbone.
I have had my small problems with DH, but as an fellow Engineer / Support Systems Tech. I do have faith in the people they use to track and squash the bug in the system and from my point of view, its unfair harrasment of the host. If you find you have so many problems, go rent a Rack Server, pay an System Engineer to handle and manage it and you have your SLA 100% (99%), but mind ya, things can still go wrong.
Why does the people that complain about critical sites and profits getting lost not have redundant websites ?? Are you any different ? Do you do the service for your clients to have redundancy put in ? Interesting questions actually….
My cents on this issue… flame away….
- Explorerdk
This is my third year with dreamhost. I’m not sure why I’m so reluctant to move off of DH considering the number of times recently I’ve been unable to access my mail, websites, etc. The first year and a half was fantastic, really no downtime at all. Maybe it’s the memory of how good DH *was* that’s kept me on board this long.
The starry-eyed DH-can-do-no-wrong, get-dedicated-hosting, etc don’t get it. DH *was* great, and lots of us remember that. It *was* a great deal for ~$20/month, but that was back when they had a reasonable degree of uptime. It’s getting embarrassing. Embarrassing to explain to clients why websites and email aren’t working, and embarrassing to read these “whoops, we fucked up again, SOOORRY!” status postings.
Gene, ok, we get it, your sites are up and running fine. You love Dreamhost and will until you die. Great. For the vast majority of the rest of us, our sites are NOT fine, and have barely functioned for almost a week. This is not opinion, this is not spoken out of anger, this is fact and can be sustantiated any way you choose. I’ve been running sites for close to 10 years and have never in my life seen or heard of any problem that could cause days of downtime on a regular basis. The only reason for this is horrible management on DH part, be it hardware related or not. We don’t need to see their facility to know it doesn’t work because it doesn’t, no matter how you want to look at it.
Even if DH stayed up every minute until the end of the year the uptime would still probably never hit 90%. Hardware fails, software crashes, but when 1500 servers and 200,000 sites go down for a week there’s a bigger problem, anyone can see that. Might want to add that to your sales page.
Brian: I didn’t say it was perfect. But I do not believe that the majority of DreamHost users are having problems. Other than the widely-advertised problems, they are a minority of a minority. People who have no problems rarely bother posting.
No, 200,000 sites did not go down for a week. Last night’s problem lasted about an hour at the most.
Peace,
Gene
My site is currently unreachable … I don’t know if this is part of last night’s problem or something new.
I hear your point, but I have a fairly large field of friends (15-20) with medium-large sites who use DH and most of us have been having problems for the past week (server stats look like combs or have random complete drops for 1-2 hours). A lot of people are posting they have problems, but the reason more aren’t is because they simply don’t know. If you get 100 visitors a day and lose 40% of them for 7 days, and don’t monitor your stats at all, you’ll probably never know. If you get 10,000 visitors a day and lose 40%, and monitor your site dozens of times a day, you’re going to notice big time. I’m not basing my judgments off just my sites like you appear to be doing. For me, it’s closer to 100 across a dozen servers. “Minority of a minority” isn’t even close to reality. To be fair, 200,000 probably isn’t true either, but statistically I’d say many, many people are being affected and they just don’t know it. And no, I have never been this affected with any other host, so this isn’t normal even for shared hosting.
I’m being moved to a new server, so we’ll see how that goes. I’m very sick of all this and will be very glad when I’m gone. Best of luck to those with working sites, I hope it sticks.
I believe the mysql server is not running right on poprock. I can contact all services but mysql. Is this problem related to the outage last night?
mi site es down
where u move ?
me want move to
To the get-a-life, mellow-out crowd: I think we can all understand that shit happens, it just seems that DH has been eating laxatives in the last year or so. Like many people, I depend on DH to keep my sites up, and I have recommended them to many friends and clients, which, as has happened to many people, is now an embarassment. I am fortunate enough that my clients are understanding and that I have not had any critical activity during this latest outage. If I lost a job because of this, I would be plenty pissed, and wouldn’t feel bad about sounding off about it.
As for the what-can-you-expect-for 7.95-a-month crowd: I understand that DH offers a lot for very little, but I would ask how many of those users are coming anywhere near their plan’s limits? This is important, because their cost model is probably based at least partly on their customers’ using a larger part of their plan than they actually are. In other words, they are probably making a tidy profit, which I don’t begrudge, but I don’t need to hear that a good deal for me means they can give me half-assed service. In any case, if I thought that a good monthly rate meant 85-90 percent uptime, I would probably be unwilling to sign up for it.
Finally, I have said this before on comment for this blog, but I think that DH’s jokey attitude does not serve them well. There is nothing more infuriating than staring at a picture of some partying hipster on DH’s home page when your site is down, or reading the Whooopsie-Daisy! explanations of outages. I work for a stuffy IT company and wouldn’t suggest for a moment that they should tighten their asses up for everything, but for Pete’s sake, it would help the customer perceieve DH as a professional outfit rather than a bunch of goof-offs.
I have been with DH for four years and the last year has been nearly intolerable. I would pull my sites were it not for the extreme about of hassle that this would be. I can only hope that they will get their act together, and in the meantime, I advise you all to backup, backup, backup, and backup your data stored on their servers.
‘As a sysadmin I must say that I run a bunch of servers in a Datacenter at my institution. Since we use a wide variety of servers we’re able to replace a downed webserver very quickly with no data loss. We control the switchs and we simply tell the switch that another machine is server that IP address. Then that machine’s webserver is told to serve the various sites. They all mount the same remote filesystem so they all have access.’
Well hurrah for you!
Now let’s engage the old cranium and have a look at some of Dreamhost’s recent failures…
* Server undergoes DDOS attack (Shocktart). Your magical switches - no help. You can re-route the data to another server all you like, but that’s just going to move the DDOS attack with it, and bring the new server down.
* Filer head unit swapped. Your magical switches - no help. Your fileserver is just as much at risk as theirs - if it fails, then all your servers will still be working, they just can’t get to the files they need.
* Major network outage. Your magical switches - no help. Because, funnily enough, it was the magical switches that failed in the first place!
You’ll also notice that two failures in specific servers that they had recently (nehi and cappuccino), they did indeed simply switch over to a new server and redirect all traffic.
So please, could you go and share the skills you learnt from ‘network administration for dummies’ somewhere else?
Well, this must be the week for downtime. If you all think these problems with DH are bad, it actually could be WORSE.
I split my shared hosting across DH and another provider so that I will have some of my websites up if one provider has downtime.
This week, for some bizarre reason, BOTH of my webhosts have been having MAJOR problems. DH has been *minimal* compared to the other one. The other host has had outages almost every day, often for hours and hours. Their support seems to be indifferent at best, they have no “hostingstatus” blog like Dreamhost does, and when they do finally get around to fixing it they hardly seem to care when I tell them I’m considering leaving.
I can’t tell you how much time I have wasted just trying to get the ATTENTION of the other hosting provider. At least with DH I know they are aware of the problem and fixing it.
I’m not sure what the hell is going on in cyberspace right now, but I have never had a week like this where my sites are going down on a daily basis (on TWO different hosting providers)…all in a week where I started a new ad campaign and put out a press release.
I’m a solopreneur and spending $200/month or more on dedicated hosting is not feasible nor should it be required. I expect some downtime once in a while. DH is at least tackling it and compared to the OTHER hosting provider (that I will not name), they are doing a much better job in responding to it.
@Stephanie: This echoes what I said. Other hosting services have problems too. They just don’t tell you exactly what’s going on, unless it covers a large portion of the network — even then you may have to search for the information.
Remember, everybody who has a Linux service is probably using similar equipment.
Peace,
Gene
Yep, the grass definitely isn’t always greener. I’ve been using paid hosting for the last 7 years now. I started off on norcap, who then became cityhosting, and now is probably something else if they even exist anymore (and I hope, for the sake of their customers, that they don’t.) My site was once down for 2 whole days, no email, no announcement, nothing. Tried calling the customer service # on their website (an international call for me) and got a “this number is disconnected” message. Finally, in complete frustration, I found their email address, and sent them a nasty letter, threatening to report them to their major affiliate, and at last got a response with profuse apologies and little more.
Then I changed to webmaster(s?).com 2 years later. Uptime was great, support was great (responses generally within an hour), but they had a nasty habit of reconfiguring their servers with no forewarning so that things (ie scripts) suddenly stopped working. Then one day my control panel got hacked. I wrote them an email about this, and was told “yes, it appears your public_html directory has been set to 777.” Well, I certainly didn’t do that. They didn’t change it back either! And they didn’t seem at all concerned about the implications of their control panel being hacked. :/
Then I went with digitalpacific about 2 years later. They were great, the only downtimes were maintenance-related, therefore scheduled and announced via email a week or two in advance. But their control panel sucked, so I looked for something more professional, and ended up here, as I had a few friends using dreamhost already. Now, I’m getting a bit tired of the downtime, so I’m looking at another host, but I don’t need to move until the end of the year. So, having found a host I’m very interested in, where everything else looks great, I’m now monitoring their site and a few of their hosted sites through a tracking service for uptime. My advice to anyone looking to leave would be to do your research very well. This is far from the best place I’ve ever hosted with, but it’s certainly not the worst either.
my site http://www.texturedjs.com is still down
There seems to be a problem with the connection to the server with all our sites… is there some major downtime happening here??
Still down.
Also:
villarisarasota.com
innostorm.com
omforhikers.com
There seems to be a major problem going on. Please update us about this problem.
@Ross&Darren and a few others; Did you try to ping the machine you’re on? If you can, then it is most likely not a network-related problem per say.
You’d benefit from gettin a monitor service from http://host-tracker.com and http://mon.itor.us - this way you can follow your site’s uptime quite well.
If you still experience downtime, then file a support ticket. (No use in posting comments like “my site is down” on this status-blog)
Cheers!
@Henrik: You know as I do that asking folks to file support tickets rather than expect help here won’t work, unfortunately ;(
Peace,
Gene
Gene - I’m pointing at my nose…
Because you are spot on with your observation…
Still database connection is not running. Control panel is dead slow and I can not wait half an hour. DreamHost has very little credits left to repeat these scenarios.
@bspiral: And you see the message right after yours?
Peace,
Gene
@Solaxim: Your problem isn’t DreamHost’s fault with the WordPress blog to which your name is linked. It’s an issue with WP-Cache. If you disable it, you must remove a link in your wp-config.php file. I did it myself once before I learned — you see I have brought my site down a couple of times too, and I don’t blame the host. I blame myself
Peace,
Gene
I don’t know where to stand on this issue (or rather, “these issues”!)
If it was for a personal blog or something, I wouldn’t be that bothered. However it is true that things are getting screwed a bit too much lately (this is my second year with DH) and there’s that thing called keeping promises, especially if people are paying for it! I’m only using ~0.008% of my allocated disk space and ~0.0008% bandwidth so it’s not that I’m being a burden on DH’s shared resources. At least semi-decent uptime is all I’m asking in return.
I’m not concentrating on freelance work for profit right now, but I noticed my only hosting client silently changed to another host just days ago, so maybe I too am losing customers, it’s just that it isn’t as critical for me.
I have two small questions to ask those more knowledgeable with site monitoring, because I didn’t bother monitoring my stats until now:
1. When “the mail is down”, is it simply a case of not being allowed to access the mail server to retrieve mail, or is any mail send to my address during that mail downtime period destined for oblivion, as if it was never sent in the first place?
2. I’m not particularly happy with the stats package offered on DH. Is there any other package that we can install (to monitor site availability and 404 errors and so on), or will I have to download the stats files and process them offline?
Thanks beforehand for the suggestions
Henrik,
I did send a support ticket.
Soon I will be on business FIOS with a static IP and a shiny new server of my very own. I may not know everything but that is just as good as, “We don’t support that.”
Darren
In the hosting world there are only a few options. Pay to let someone handle your hosting needs or do it yourself. Relying on someone else to host your content has many benefits considering the way the services are bundled and offered to us through simple and easy-to-use control panels - this is the reason i’ve chosen DH. They bundle it up nicely - and it *almost always* just works.
The second option is great for those who want to use their own desktop machines as servers - for mail or http - or whatever, really. Its great if you want to throw up some music or images or some burned dvd’s - but then we get into running your own DNS servers. Now - if we aren’t using core server platforms - most of seem to be bent on the most cost effective means to propogate our DNS and pair it to a domain name. Most times we are limited to our local ISP bandwidth (and up/down times) to keep our sites and projects going. And then we need our array of server side wares including PHP and ASP and Python environments. It’s all great for the guy who wants to play for himself and develop on his own local server environment..
But now you’re in the league of maintaining your own active/live server and this has a habit of consuming a person. Been there done that got the tshirt. One, even 10 sites might co-exist very well on your locally run Server machine depending on the usage and bandwidth requirements.. But is this the best option for your clients?
From a personal perspective it is time consuming and nerve-racking maintaining an array of servers - whether they be boxes or racks or unix or windows. Servers are servers. They require love and attention. We can’t all love back. (well i can’t :P)
I keep on my desktop the ‘Server Ping’ gadget - whereby I have the IP’s of my sql, http, ftp, smtp etc. accounts listed - with a ping refresh every 60 seconds on each IP. This is a good way to keep up-to-date with your own IP status, DH account or not.
Long Live DH!
–
DreamHost gives you great bandwidth and storage, they just don’t give you enough memory and processor to use it. If you have a site that no one is going to then it’s great. If you have a popular site forget it, not to mention DreamHost is up and down like an elevator.
Long Live FIOS and Unbuntu Server!
Dreamhost’s service isn’t perfect, but it’s good enough for the money we pay. If your business is critical, and you can’t allow a certain amount of downtime, you should have a backup plan, a secondary webhoster with better uptime, perhaps?
I don’t use 2 webhosters. I use 5. And that has a reason.
I’ve used and reviewed many webhosting services, and if I still am with Dreamhost, is because they deserve it. And I renew it in a yearly basis!
Just a few comments:
- I purchased DH shared hosting service recently, so don’t have any real impression of their reliability, but price per amount of storage and bandwidth is real bargain. I plan to use it for my and my friends’ photo galleries, a wiki or two, few mailing lists and such. It’s probably adequate for that kind of use.
- If DH gives SLA-style guarantees, I’ve missed them. It *would* be nice if they would do that.
- There are some less that ideal details I have noticed: number of pre-configured (’one-click’) apps is rather limited, support wiki is a bit capricious (loosing context info, logging me out without provocation….), rules on background processes are rather vague…
- OTOH, shell access, use of cron etc are rather rare features at this price level
- All things considered, I would not use DH to host my serious clients’ sites. There are providers with fault tolerant clustered servers, SAN storage, multiply redundant network interfaces, switches, routers and uplinks, on-line database backups, set-in-stone SLA commitments etc. They come at a cost; say, starting at order of magnitude more expensive than DH and up.
Providers like DH have their uses. Those 100 times more expensive have theirs. Pick according to your needs and budget.
I’ve been using DH for over a year now and this little incident is nothing compared to what happened over the summer. So I don’t see why people are complaining so much.
1. Hardware will fail, it’s inevitable.
2. If you are hosting mission critical applications on shared hosting, it’s your fault, not DH’s. Invest in a small server or Virtual Private Server.
3. The problem was resolved within an hour, so what’s the fuss? An hour of downtime isn’t going to kill everyone, and if you really rely on your site that much, refer to #3.
4. Price point, you’re probably paying $7.95/mth for a ton of features that some other hosts would charge $24.95+/mth for. So expect hiccups.
5. Human factor, we all make mistakes, big or small, it’s human. No one or thing is 100% perfect.
6. Straight to the pointness, DH pointed out the problem and fixed it. End of story, no more bitching and continue on with your happy little life.
Honestly people, no one host is perfect and I know, I’ve been through a dozen hosts in the last 5 years and at different price points they all had their share of problems. I admit that DH has been shaky and that I’d like to see better uptime, but the support is probably one of the more human ones I’ve seen to date and I can’t argue about price either. My last host was charging me $20.00/mth for 5gb of space, 50gb of bandwidth and 10 MySQL databases, so I’ll take the occasional bump in the road to keep what I have now.
Lastly, we were notified that a server move was happening and that there would be kinks, it’s up to you to act accordingly with the move (aka. move mission critical applications to a different place until things are settled in and working).
Murphys Law says anyone (let alone decent people) have things go wrong at the worst possible time. In this case the worst possible time is ‘always’ because the law ‘you can’t please everyone’ comes into play.
Most people faced with a person who has neither arms nor legs thinks ‘oh you poor soul’ at first thought. Realise what you’ve got and how well it works for you, as well as accepting that life will never be perfect.
I’m quite pleased with what I’ve been given since I signed up with DH. As has been said above, if it’s that critical perhaps spend your money on hosts that aren’t affected by such horrible things as the facts of life.
Enjoy.
You screwed up my career Dreamhost! I was in line to be Attorney General Alberto Gonzalez’s replacement when he’s fired/resigned but when Bush and Karl Rove came to my “Mittens for Your Cat” website it was down! Now I’ve been eliminated from consideration!! Thanks, Dreamhost….I need 110% uptime for important clients like this. I expect the service to be up 24 and 1/2 hours a day.
Anyway, its all your fault my life sucks now. I was next in line to be Attorney General, now I’m living in a washing machine box under a freeway overpass. Thanks Dreamhost!
How is it that a single 10GigE interface failure can take out half of the entire set of servers? Any rational data center network design would have dual interfaces at critical networking points, to avoid this kind of failure mode. 10GigE ports on routers and switches aren’t cheap, but neither is the cost of losing half of your servers and having your service fail for large numbers of customers.
Critical connections supporting 50 percent of all servers shouldn’t go down due to single interface failures. There should be multiple interfaces on those connections configured as a port channel Ethernet, or as multiple L3 routed paths, or whatever works best depending on your network design and failover preferences.
holy shit… is dreamhost always has downtime everyday?
Well you see, I had my entire future and career fully dependent on the fact that my $9 a month website would be up and running flawlessly for a specific, precise 3 hour window. It certainly has to be someone elses fault that poo happens… it can’t be my responsibility to make sure I have a way of accessing my site locally, can it? Nope, I’m just an innocent victim here… oh woe is me. Oh, and I was ALSO in line for attorney general. Go figure, two candidates and we both use dreamhost. Jim, any room in that box?
Wow theres some little whiney bitches on here. Shut up they are doing what they can.. if you think you can do better then why the hell arent you in their position? Shit Happens.
As for all you who say, oh I have 50% uptime. No you dont, if you have 50% uptime you wouldnt be here bitching. Use host-tracker.com See your uptime for real. I have a 99.17% uptime, including this outage.
@Tom: 99.72% here since the full outage in February to fix the faulty power cable.
50%, right.
Peace,
Gene
@Chas: good point, a design must have no single point of failure
for the rest of you please remember that we are not all cheap bastards like you and spend large dollars per month on multiple strictly business and dedicated servers so it can be a big deal. Every DC will have problems but the common theme here is single points of failure. But at least Dh has stopped trying to get every pissant home website on there servers with a 1.21gigwatts of bandwidth and server space per month and seem to be looking at getting a different structure in place. Also response to help desk emails is also getting better but a long way to go
Tom, you must be a retard. The box may be on but there is usually no http access. So lol.
i have respect for “built from the ground up” companies, if DH were some cookie cutter rig that was built by a monster corp i’d expect 90% up time (and allot of morons at the helm). I’ll take a pack of wild geeks with purpose any day. I think the great attitude and willingness to take responsibility for there imperfections more than makes up for the few flaws.
these are a few things i appreciate about DH:
1.never had a billing error
2.never had to wait more than an hour for an answer (even to stupid questions that i sould have found for myself by reading a prominently posted faq)
3.an amusing newsletter
4.genuinely pleasant folk
5.they haven’t become assimilated by google. yet. (at least hold out for big money guys, if you sell out at least make it worth it)
6.prominently posted faqs
7.based on the pictures i’ve seen they are, on average, not much better looking than me.
i should prolly be working…
thanks folks, you provide everything i need from a host.
bishop
@Josh, actually no. It uses an HTTP GET, POST, or HEAD, to see if HTTP!!!! functions. Why dont you try the stuff before you attack me.
It’s very annoying seeing the amount of people acting like little kids on here. In terms of value for money DH offers excellent service, and I am still very grateful for the way they always update about what’s going on, and get back to you pretty quickly if questioned.
Someone above me who said that it shouldn’t matter if you are paying $5 or $50, the service should be the same, well of course you know it’s not going to be like that, surely.
Well I’ve just found this thread via the latest DH newsletter, and have read all your comments without actually going into a coma. OK folks I admit it after reading the comments on the newsletter about ostracising adult sites onto servers just for them, because they are causing all the problems. (Gotta find a scapegoat eh DH.)
The downtime was all my fault. Wicked me and wicked adult sites!! At a guess the crashes and downtime were probaby caused by the 3 geriatric males viewing my topless photos. Those guys take ages poring over the details and slow up the servers. Slap my (__/__) and call me wicked!! But don’t blame DH.
Actually I’m willing to forgive DH for trying to pass the buck onto adult sites. I think DH do an excellent job and I have been with them for 4 or 5 years now and host 4 sites with them. Yes this year has been a bit problem ridden compared to the last few years but I consider a little downtime to be beneficial for my 3 geriatric male surfers because if the server didnt crash occasionally those guys would go blind staring at my photos.
Every cloud has a silver lining.
Give the guys at DH a break folks, theyre just being kind to the elderly.
Best wishes to all.
Leggy.
Elemanyak.net still down.
Since I joined DH 2 weeks ago (exactly 2 weeks), there has been 3 or 4 times that the site was either slow or was simple down. My sites are still not stable.
I just hope it will become beter in the future.
Get used to downtimes, but also get used to low prices
200GB for 100$/year…
I just hope it will become beter in the future.
I wish that
Well I’ve just found this thread via the latest DH newsletter, and have read all your comments without actually going into a coma.
thanks
Thank you for informations
Great job done! Dreamhost is the best hosting you can find!
TOO MUCH CPU
On your favorite shared web hosting, say we put up a website like youtube.
Upload a single, one and only, 20mb video.
Let’s disregard uploads.
QUESTION (without abusing or “using too much much cpu”)
1) what web hosting?
2) how many people can view the video at the same time?
3) how many people can view the video each month?
i know that dedicated hosting is more suitable, but
is there any shared web hosting where we can put that kind of website without running into any “too much cpu”, “abuse”, or other similar problems.
- slimetoner @ yahoo.com
Hey, don’t think us dedicated people get better service - in fact we get worse. I wish we never got a dedicated DH server, let alone two. The panel only slightly works with the DH Dedicated solution, and downtime is much higher on those servers than our shared accounts.
There seems to be a problem with the connection to the server with all our sites… is there some major downtime happening here??
This was the one day I couldn’t deal with outages like this. I had major clients coming to my site and now I may have lost all hope of building my client base.