If your server is down, how long can you go without a reply from your server people?
 

If your server is down, how long can you go without a reply from your server people?

how long til 1. you start getting angry?

2. you start thinking about moving hosting?

3. you will move your hosting asap?


it's what, 2:14US central time, and i cant get a reply from icq / suppor ticket / email / private message

this is fucking bullshit, :disgust

By the time I contact them, they should know its down already and be working on fixing it. If they find out its down from me, that's very bad on their part and probably a host I won't work with for long.
WG

Exactly....

But....For what it's worth.....I've never had to wait longer than 30 seconds to get a response from ISPrime.

our hosts tech's get a page if the server goes down, so they know way before I do, ive never had a problem with either of them

www.mojohsot.com
&
www.webair.com

In all fairness, there are situations where the server may appear to be down, when in fact it is not, but circumstances make it appear to the end user that it is down.

You shouldnt have to contact a host about a crashed server, they should already know.
In the off chance you have to contact them, I would expect to have some communication no longer than 15 mins.

DDOS from that ex-GFY member

10 seconds... I have NO patience with stuff like that. Been happy so far.

Same with Phatservers for me. :thumbsup

JScott...You need a change :)

WTF?



.....

The host can ping my IP's to tell if the server is up and they all have remote logins to make sure apache is running. What other circumstance is there where it may be down but not give the appearance of that?
WG

:thumbsup



Problems with the isp, dns issues with the domain register.. There is a number of things..











jscott- Good luck getting back online and moving to a new host :thumbsup

A DNS issue would likely mean that all their customers are down, I would hope they would notice something like that. If a registrar is having a problem, that's not related to my host really.
WG

they know its down b4 i do , i call and they say "Hi we working it , go chill and get some pussy"

Wiresix.com ??????????????/

I think Bagdog meant if a part of the route is down, so the client can't reach the server .. So up for some, down for others.

I use Uppanel to monitor my boxes ( ftp, http, mail, bind ) so I normally know if a box is down.

If your register has a problem resolving your dns for your domains then you are correct, it is not related to your host, but it will cause you to not be able to excess your server via the domain :winkwink:

massive FIBER gives me a hardon

I would say 10-20 min max

Ok, I have server contact now at least :)

BUT, problem still not solved :helpme

an hour to make contact? Wow!

Did you try calling them?

remember, i was angry when i posted this thread, i had already waited hours before posting this, i was NOT angry the 1st few minutes of my server problems

Thank you

Im still just finishing a very small hosting with a private host (he runs like 4 servers out of his office) and any scheduled time is warned about a week or so in advance EVERYTIME and when server goes down (it did once because one of the box's blew a stick of ram) he contacted me and told me b4 i even noticed!!! lol

My latest web host problem took the piss as getting a reply from him took about 12 hours. Thats why I cancelled in his 7 day refund time and now im claiming back with Paypal :)

My new host, ive paid more bcos i want good service :)

i use layered tech and they worked well for me.

My servers don't go down much but in the event they do, my host knows it instantally before I do, and it's fixed asap way before I even contact them.....:winkwink:

LOL. That's some service. It would be even better if they sent over a few complimentary girls to keep you busy during the downtime.

hey guys, not talking about servers "going down" talking about DDOS attacking making them, forcing them to go down

Have you tried calling them?

You shouldnt have to wait at all. Your host should already know there's a problem even before you do.. With all the support options available, web, phone, email, it should never be a prob getting in touch. I bet they never have problems reaching you when your server bill is due ;)

I'd find another host...

thats the ULTIMATE package..... it costs too much 2 justify the price :pimp

then you are pretty mch fucked man if your getting DDOS'd

i would reverse DDOS em back

I don't know, if the end user can't access my server then I'd consider it to be down...

About an hour - they usually know there is a problem already and working on it and usually reply within 20 mins.

Around six hours - I'm tolerant :) Depends what the problem is and little point in being pissed.

Already preparing to move by the end of day two and usually have a host lined up - but, again - depends on the problem and communication.

On day three the move is well underway and prob operational on a different host by the end of day three (usually got a lot of files/scripts to handle).

Little point in getting pissed - it's frustrating, but there can be all kinds of reasons - just stay cool and start the clock. If there is some communication, that's a good start - but watch for the bullshit excuses :thumbsup

Different story if that is genuine - and may take time to resolve and pipe a load of crap back in response to the DDOS stuff.

Can't really blame your host if someone is attacking your server... no matter what host you switch to, you'll still be down for as long as he wants to keep attacking you.

If you upgrade your server hardware to something that can handle the load that may fix the problem, but it really depends on what kind of resources he has to hit you with.

Six hours before getting pissed... damn man, you are tolerant. I'd be right pissed off around the 20 minute mark, downtime is expensive.

webair is always a call away... I have Gerard's private number to his home for these matters and for phone sex as well.

Their live online support is always around as well

:1orglaugh I've seen the "other side" as well stev0 and if the host has real problems and it's only a matter of time spent working thru em - no biggie. Most stuff can usually be remedied in minutes tho.

Avoid this scenario like the plague, but don't have more than one server with any host - then it would be literally a total collapse.

The only real problems ever with hosts have been with the larger hosting companies with "staff structures" and where there is foul communication between them. The worst ever was with Rackspace where everyone from the CEO downwards talked shit and never kept to what they promised - tho, all would have been fine if the "staff" stayed out of it and left the job to tech support who were actually very good.

I get pissed within a few minutes.

What if I can access it, JScott can access it, Webby can access it, but you can't access it? Is it still down?

well thanks guys, apparently the attack has stopped (for now)

my server guy finally did reply to me, and was doing what he could to get my sites up.........it was my sites being down that led me to be angry with anyone involved.......me, my server people, and very much at the attacker

but, nothing can do about it now, he's done what he's done, now time to TRY to recover.........he won, i lost...........he's smart, i'm stupid..............he now has me by the balls (along with others aswell)

If you're down due to DDoS that's a tricky one. addressing a couple things people have spouted about here.

#1. "Hard Down" outages are easy. If apache fully crashes or the machine goes down, any host can trivially see this sort of outage. This outage should be extremely rare, as uh.. contrary to public belief this type of outage doesn't actually happen much.

#2. "Soft Down" situations are much harder to setup automated monitoring on. What defines "down"? Apache might be serving pages great, but the wrong content (e.g. mis-configuration) - or any myriad of other items where it truly appears data is being served, and not triggering alarms. This is where the host needs conent-based monitoring capabilities. This is NOT invisible to you, you WILL have to work with the host to set them up.

#3. It's absolutely impossible to monitor for ALL outage scenarios. A good host should be monitoring for everything realistically they are able to, and coming up with monitoring solutions for you when something is found that previously was not caught. For these situations, 24x7 support (via phone!) is critical. A "server down" call that is actually a server down, should get an immediate answer on the phone and then escalation right from there. Generally you should know what's going on within 15 minutes of your call. E-mail is the same way, but I advise against it as it's a somewhat unreliable mechanism. AIM/live support/etc. may be an option depending on the host - personally I hate these options as it makes support on the host's side a real nightmare (what did tech A say to customer, do, etc. when there is no ticket involved?). Depends on your host's support options, but you should be getting someone looking at stuff pretty quickly :)

#4. DDoS is hard, lets go shopping! For folks being DDoS'ed, throw your time estimates out the window. For most attacks, it's probably pretty trivial to filter and you'll either not even know you've been attacked, or be down for an hour or two until your host can put some filters in place. However for the REAL attacks (e.g. multi-gigabit, millions of packets per second, non-spoofed traffic, etc.) don't expect much from almost all hosts out there. These attacks are EXPENSIVE to filter, no matter what anyone tells you. First the host needs the inbound capacity to begin with (routers that can actually handle 50 million packets per second are not cheap), and either some very expensive DDoS filtering equipment or a lot of spare hardware and some good experienced tech's to deal with it. We've had customers under 4gbit+ attacks for over a month, where it took an entire rack of dual xeons to filter out the bad traffic. I feel bad for the clients, but this does not come cheap - someone has to pay for the equipment investment and the few gbit of bandwidth being burned.

If you are truly being attacked by a "real" DDoS attack, you need to find a host used to dealing with that sort of thing. The alternative is buying filtering services from a DDoS protection company such as prolexic. Unfortunately neither is a cheap option. However there is good news! If you are with one of the major "adult" hosts (as in, top 5) most of these guys have dealt with DDoS before and at least know where to start. Some are better than others, but you are infinitely better off with folks that know wtf they are doing, vs. mainstream reseller-flavor-of-the-month who will not have an idea of where to start. This sucks for small hosts out there - to be able to quickly adapt to these requirements you need oodles of spare capacity and hardware readily available, which is cost-prohibitive even to hosts of our size.

In short, good luck to ya! communication is key when you're having issues, and I have to say as a host it's easy to forget this in the heat of the moment. Generally I'm furiously working on an issue when the customer ticket comes in, and tend to reply after I get it figured out or the site back up. This sort of tunnel vision just leads to upset clients in the end, even if you did get their stuff back up faster than the 2-3 minutes it would have taken to respond. Live and learn :)

-Phil

Great post man. Reflectednet just got a plus in my book :thumbsup . However, on the other side, i very much doubt he's having issues of this kind, as getting the botnet of that size is not as easy as most people think, and to hire someone that operates a botnet of that size would be very expensive for that other person aswell, actually alot more expensive then it would be for him to buy DDoS protection equipement.
Most DDoS's that i've seen have been just few rooted machines running programs that saturate traffic, sometimes try to spoof it through socks (which also cost the person doing DDoS as these are special socks that can handle the bandwidth he's sending through them), or some other form of attack where they are overflooding the apache with requests to database, or flooding some application running on the server that makes the apache choke, or overloads the server.

yeah - must admit that i'm terrible with down time too....

You all should have a look at Neustar's DNS network to ehanced your redundancy, expand to a global infrastructure, and ehanced security mitigating against DDOS, Cache Poisening, Pharming, Port Scans etc. Real-time propagation and much more. Let me know if you want to talk. Also have lots of connections in the hosting arena I can refer you to.

icq: 428-946-566