Central Services and Spunky Cluster Move
Due to continued space and power constraints in our primary data center, we will be moving all central DreamHost functions as well as the “spunky” cluster to one of our newer data centers. This move will begin Friday, March 21st, at 9PM PDT, and is expected to last up to 12 hours, until Saturday, March 22nd, 9AM PDT.
Central services include panel.dreamhost.com, dreamhost.com, dreambook.com, files.dreamhost.com, media.dreamhost.com, wiki.dreamhost.com, discussion.dreamhost.com, basically everything but secondary DNS and this status blog. Primary and tertiary DNS will be going down as well, but will have a more limited downtime window of only 2-3 hours. For spunky, all web servers, mail servers, file servers, and MySQL servers in the spunky cluster will be unreachable. The list of affected spunky web servers is here:
abbazaba absinthe achilles adams admiralty aero alameda almalibre alondra altoid alvarado archibald arrow atlantic avalon aviation babyruth badillo balu bandini barranca barry baseline bazooka beaudry bellflower beverly bixel blowpop bonbon bottlecap bounty bourbon boyd broadway brownie bubbleyum bundy burton butler butterfinger cahuenga caramello casamento catalina centinela century cerritos cheeto cheezit chex chiclet cienega cognac colbert compton cookie cordial crackerjack crenshaw crunch cupcake dingdong doheny dorito drumstick dumdum edendale euclid fairfax fc23 federal fig fireballs firestone fletcher flower foothill formosa frito frontseat fruitland fudge fundip garland garth garvey gobstopper goober good gower grand granola grappa gumdrop gummi haven heath heaven highland hillhurst hillmont hoho holt hoover hope hops hyperion jalapa jawbreaker jefferson jerky juniormint jurupa kidsgames kiki kitkat krackel lankershim larchmont laurel lemonhead licorice lifesaver limbo-spunky3 limbo-spunky4 locust look lucas malt manchester mariposa mayo melrose mento milliken mills monroe moorpark mounds munchies nash necco nerds nilla nougat nutrageous oakdale ocean olive olympic ord oreo ouzo overland oxnard paramount payday pernod pez pico planters plenty pocky popcorn poprock powerbar pretzel price pringle rampart razzle redhot redondo reeses rexford ritz rockyroad rodeo rollo roscoe rosecrans rossmore rum runts sake saltine sanborn sawtelle schnapps scotch selma sepulveda shocktart skittle skor skybar slauson smores snicker snocap soto sourpatch spree starburst sugardaddy sunset taffy tampa temple tequila toblerone toffee tootsie topanga torrance truffle twinkie twix twizzler union vanowen vermouth vine virgil warhead washington western whittier whopper wilmington wilshire york zagnut zuma
We realize this is a very large move and we will have a dozen techs onhand to supervise. My apologies for not posting this earlier.
Happy DreamHost Mover Team!
.
March 12th, 2008 at 10:44 am
Wow, I feel really sorry for all those on Spunky. Good luck to the DreamHost Uber Techs.
March 12th, 2008 at 10:45 am
Ohh this is a huge 12 hour downtime!!
Man this will surely have 500 comments I tell you :p
March 12th, 2008 at 10:45 am
Good Luck! My sites still remain up and fast!
March 12th, 2008 at 11:25 am
I’m on twix. Will you give me a twix if my server gets dropped and all the hard-drive platters get offset? I’m hungry. Thanks.
March 12th, 2008 at 11:43 am
“For spunky, all web servers, mail servers, file servers, and MySQL servers in the randy cluster will be unreachable.”
Hold up… you are taking those of us on Randy out yet again?? http://www.dreamhoststatus.com/2008/02/08/randy-move-going-ahead/
March 12th, 2008 at 11:54 am
So it looks like a side effect of the move will be a roll-back from daylight savings to standard time… heavy stuff …
March 12th, 2008 at 12:00 pm
Will this affect email as well?
March 12th, 2008 at 12:02 pm
“Happy DreamHost Mover Team!”
You guys might be happy, but I doubt anyone on the affected servers will be.
March 12th, 2008 at 12:02 pm
Looks like DH is pulling a Spitzer -
Wonder how many clusters there are
March 12th, 2008 at 12:03 pm
The only reason why I saw this is because my SMTP server just told me to go piss off. This is getting ridiculous. It is always something. I am really starting to consider switching over to those other guys (*cough* Media Temple *cough*). For the past week my email server on Spunky has been kicking my ass up and down the digital pipeline. Snail mail is beating my electronic mail to my clients. I know that is an exaggeration, but not by much.
March 12th, 2008 at 12:29 pm
One might think - maybe spunky should be split into two or three clusters so the next move doesnt affect so many users, and so it doesnt take 12 hours to move all those servers!
March 12th, 2008 at 12:51 pm
Spunky MySql servers too? not that it’d make a difference if they were different places..
I’m in Spunky, never really had any problems other than the major stuff.. I just don’t like that it’s easter weekend.. lot of people home.. (not me though..)
Good Luck!
Thanks!
March 12th, 2008 at 12:52 pm
I’m a Spunkian so I expect my service will be affected. Obviously the situation isn’t ideal, but I’m not going to complain about it because I’ve enjoyed a marvelous level of service over the last few years and I accept these things happen from time to time - particularly in a shared server environment like this one.
March 12th, 2008 at 1:04 pm
This comment may not be “official” but you officially suck.
March 12th, 2008 at 1:18 pm
My email is down already …
So 12 hours … Good luck.
March 12th, 2008 at 1:27 pm
Email is up! Email is down! All freakin’ day…. I mean week. Come on guys!
March 12th, 2008 at 1:43 pm
I’m really sick of having websites randomly have to move, and my e-mail not working 75% of the time is pissing me off.. I’ve been through your billing glitch, the unexplained outages, but this isn’t even funny anymore!
March 12th, 2008 at 2:49 pm
I think the Dreamhost CEO is partying too much instead of getting his business run reliably … http://flickr.com/photos/thingee/2322196422/
March 12th, 2008 at 2:51 pm
Thank you Dreamhost. This on top of the four hour downtime last night is just what I needed to motivate me to finally make the transition to another host server. I’ll be making the move within the next week…that is if you can keep the system up long enough for us to make the change over….
March 12th, 2008 at 3:09 pm
Yes, because screaming and whining about it is gonna help. If you’re trying to make money off of shared hosting, then you can blame your own cheap nature or bad budgeting.
March 12th, 2008 at 3:15 pm
I have my server (chiclet) overloaded for months and DH does not solve it in any way… and now, 12hours of free downtime??.
NO thanks.
I think that DH is in the way to the DEAD
The unhappy DH customer.
March 12th, 2008 at 4:14 pm
@Nicole — Thanks for pointing that out. Erroneous copying of that line from a previous post on my part. Fixed!
@P — Dar. I remembered to change the first PST but not the second. Fixed now. I did get someone to proof-read this too…. I swear.
@OnyxRaven — Yes, spunky is in need of some splitting.
@Everyone else — We are aware of the inconvenience this causes and we’ll do everything to minimize it. If it helps at all just know that a 12 hour downtime window for you guys typically means a 16-20 hour day for the Admins. We’ll try to make it as painless as possible. Check back on the post next week for updates.
March 12th, 2008 at 4:59 pm
Gobstopper noooo!!! Just jokes, this is the first time that I can recall my server having scheduled maintenances. I stay for the features, you guys are hard to beat in that regard.
EVERYONE ELSE WHO IS BITCHING AND COMPLAINING ABOUT LOSS OF SALES
Take the time between now and the move to setup a SECONDARY ACCOUNT SOMEWHERE ELSE, COPY YOUR SITES, AND WAIT TILL EVERYTHING IS MOVED.
The point of these ADVANCE WARNINGS are so that you can do something about it your own damned lazy selves to mitigate your losses.
March 12th, 2008 at 5:14 pm
how do I find out what server I am on? Im on a vps
March 12th, 2008 at 5:27 pm
@bob cobb: Log into the control panel then click on “Account Status” at the top right. Give it a moment to pop down and you’ll get all your server info.
March 12th, 2008 at 6:23 pm
on a lighter note, i bet it’s a fun job trying to name all those servers
March 12th, 2008 at 7:19 pm
Thanks Micah!
make sure someone supplies pizza! (just not near the servers
)
March 12th, 2008 at 7:24 pm
dumb question of the day… What happens to email sent to boxes on those servers during the downtime?
March 12th, 2008 at 9:27 pm
MY SITE IS DOWN AGAIN! how hard is it to sort thsi shit out!
March 12th, 2008 at 9:28 pm
@Joe: it will be delayed but it will eventually arrive if the sender has a reliable ISP (that is, you may lose some spam or other mass mailings but not real mail)
(non-definitive answer from another user)
March 12th, 2008 at 9:30 pm
And the crying begins a week before anything even happens.
March 12th, 2008 at 9:33 pm
OOMG,,,,,,Not again!!!!!!!!!!!!11111
March 12th, 2008 at 9:48 pm
I’m on bowser (not on the list) and it’s down.
March 12th, 2008 at 9:57 pm
March 21st at midnight EST? Wonder why Pacman is grinding to a sympathy halt now? It usually waits until at least the day after… Must be low on power balls.
March 12th, 2008 at 10:03 pm
Is bowser down?
March 13th, 2008 at 12:08 am
How do I know what cluster I am on?
March 13th, 2008 at 12:51 am
god! Please quick!!!!!
March 13th, 2008 at 1:10 am
Man, huge move, good luck, hope it all runs as smooth as possible, looks like I’ll be planning a day outside (god forbid) for that day seeing as it’s my 10am-10pm.
March 13th, 2008 at 1:35 am
Can you give me the address for where I should send you a bill for a day’s worth of work lost because of your unannounced 12 hour downtime?
GIVE ME A FU*&ING BREAK!!!
March 13th, 2008 at 1:39 am
Oh, and while I am at it, can I send you a bill for the amount of money I would have made, if it weren’t for the client, work and referral I lost because of your unannounced 12 hour downtime???
NOT an exaggeration…
March 13th, 2008 at 3:49 am
Hey, loser, a weeks advance notice isn’t quite “unannounced.”
The only reason you’re losing money is because you’re incompetent fool. If your site was important in any way, it wouldn’t rely 100% on a single cheap shared hosting account to stay up.
On a brighter note, any clients you’re losing are better off.
March 13th, 2008 at 4:48 am
nice reply shawn.
One great thing about dreamhost is they enable us to have a good laugh at the part of the world’s population who have an IQ less than their age.
Everyone who complains here is obviously an idiot because complaining here isn’t going to help you one bit! But thanks, provides me with a good giggle.
March 13th, 2008 at 5:21 am
Is it possible, in the world of hosting, for DH to redirect down sites to a page saying, “This site is down due to scheduled maintenance, and will be back shortly.”? That’s a lot better to end users/customers than just plain broken.
March 13th, 2008 at 5:56 am
Shawn that’s rude actually. I don’t think its unreasonable to have a website that’s up more often then it’s down. While I’m well aware of what you get for your $10 bucks a month from dreamhost it’s not what a lot of people expect. I’m still with dreamhost because it’s cheep and they give you a ton of space. Not because they’re reliable (they’re not) or because they’re fast (they’re not). But new users end up feeling duped.
March 13th, 2008 at 6:36 am
@Nyhm - cool idea - I’ve got a possible solution. Maybe some of the experts out there can help out by picking this apart: If I set up my own custom “site down for maintenance” page on a different server not at DH, and set my 4th DNS server to point there, then magical web routing should get my custom “site down” page at my other host whenever my DH account is unreachable.
March 13th, 2008 at 6:46 am
@Myself - nvm - that will only work if DNS 1, 2, and 3 are down…
March 13th, 2008 at 7:26 am
@Bryan - I think you’re right; using a 4th DNS would only work if the first 3 were down.
To make things simpler, I’m not even trying to support an automatic failure mode. I’d be happy if they could do this “manually” for known service outages. This would apply to things like this weekend, and when DH staff discovers that a server is dead, while techs are fixing it.
DNS might be one approach. However, DH does not manage DNS for every customer. Also, DNS has a propagation delay, which would not be suitable for short outages.
So, it would probably have to be something they put within their network to feed a generic page for every HTTP request. Notice that this does not fix HTTPS, SSH, FTP, Jabber, etc. It might not be worth the effort at all.
Soapbox: I find it disappointing that the Internet was founded on the enlightened idea of distributed reliability, but Web services are generally single points of failure.
March 13th, 2008 at 7:34 am
Migrated MX over to googlemail and been anxious about losing mail
-I see dozens of spam msgs every hour which tells me the MX is working
-no one has ever said -hey did you get my email? and I havent had to ask anyone either
@Bryan -great idea -I’m gonna have google site setup for that as a fail safe
March 13th, 2008 at 9:12 am
Hey “pissed” - you are a complete twit! There is nothing “unannounced” about downtime you are told about a week in advance, and DreamHost’s address is easy to find, That said, a glance at the TOS to which you agreed should make it obvious that DreamHost is *not* liable to you for any of your “lost earnings” anyway.
You need to visit http://fuckdreamhost.com where you can revel in your own misery!
March 13th, 2008 at 10:26 am
mahaha wonderful usage of a domain RighteousIndignation
i needed a laugh today lol