DOS Attack, Grrr....

[piper]

How many of you are affected by the current DOS attack and what server are you on?

I'm on chamaleon and I have 3 sites down.

[Connie]

Actually I have no idea. I did receive the following e-mail from Jag.

We are contacting you to inform you about a
curennt issue at our Atlanta facility. At this
time network enginners are working to disable a
large DOS attack which is underway and preventing
proper network connectivity from most locations.
We expect to have the situation under control
shortly.

[anigmad]

i'm on admiral and being affected..lotsa complaints from my customers.

I hope it's resolved soon...they keep saying 'shortly'.

[ehost]

I'm on admiral as well and the odd thing is, one ip wont respond, but another on the same vps will.

[dreamerli]

I'm on Casper. Been down for nearly 10 hrs. Am getting complaints from clients. Been with Jag for about 1 month and find this outage inexcusable. I've hosted with other companies (shared hosting at a fraction of the cost of jag's semi dedicated) over the past few years and have NEVER experienced this type of outage (DOS or otherwise). Moved to Jaguar base on reviews I read on webhostingtalk.com. Will be leaving soon and providing my own review on webhostingtalk.com. I am VERY dissapointed.

BTW, I thought all server in datacenter in texas (says so on jag website). Altlanta is a surprise to me. Anybody know about details of the Atlanta datacenter?

[piper]

I'm on admiral as well and the odd thing is, one ip wont respond, but another on the same vps will.

I've noticed that too...

BTW, I thought all server in datacenter in texas (says so on jag website). Altlanta is a surprise to me. Anybody know about details of the Atlanta datacenter?

This surprised me too. I thought everything was in Houston.

[Julian Muñoz]

My dedicated box is on atlanta and four of my five IPs are being affected... I was lucky because my primary IP is the only one working, so I redirected all my traffic through that IP.

[piper]

I have been trying to get support to please reroute me through the working IP. So far no luck...

[rainboy]

My server was affected by the dos-attack as well, however the server kept reachable for me, only problems i had was that the server could not reach some external links in the US itself, towards europe no problems.

I can not blame JPC for this at all, this is a backbone provider problem, not a server issue. Instead blame the idiots trying to dos a server(s).

Kindest regards,
Patrick

[Highway of Life]

I'm on Casper. Been down for nearly 10 hrs. Am getting complaints from clients. Been with Jag for about 1 month and find this outage inexcusable. I've hosted with other companies (shared hosting at a fraction of the cost of jag's semi dedicated) over the past few years and have NEVER experienced this type of outage (DOS or otherwise). Moved to Jaguar base on reviews I read on webhostingtalk.com. Will be leaving soon and providing my own review on webhostingtalk.com. I am VERY dissapointed.

BTW, I thought all server in datacenter in texas (says so on jag website). Altlanta is a surprise to me. Anybody know about details of the Atlanta datacenter?

Ditto, same boat... only this one floats on katarn.

I switched to JPC based on reviews from WHT, and the last month I had 31 migraines... thought everything was solved two days ago only to hear about the DOS attack that was supposed to be taken care of "shortly"... that was 2pm.
Almost 12 hours later, my customers are asking us to switch to another hosting provider.
I guess too much downtime wears on people.

[rainboy]

Highway of life,

Yep, if you had more downtime this month, i can imagine your users are asking for it. But do not believe that ANY other hosting company can not have a DOS attack. I do not know why the other outtages did exist, and if they where related to any technical issues on JPC side, guess thats where you have to base any decisions on, making a decision on a DOS attack is silly.

Kind Regards,
Patrick

[Highway of Life]

No, the DOS attack is sort of the "last-straw".
If we had been up 99% for the entire month and then the DOS attack, then I know the customers would understand, and be willing to wait it out.
But I have had nothing but open tickets and aspirin pills since switching to JPC 7 weeks ago. I figured in the first 2 weeks of downtime to be "server-setup-growing-pains", but then 5 weeks of frustration has just made this DOS attack unbearable.

A lot of the other outages were due to the Backup Node causing high load problems. So high that it would knock our sever out for almost an entire day every time. And every single night we would go offline.

We were constantly experiencing load issues and CPU % up into the black, but everytime we reported a ticket, they would report back that the server was fine. This happened time and time again.
Finally, a friend tipped me off to check it on the 30 minute mark every half hour.
Found the IOwait reaching 100%

It turned out to be the CP was running regular crons every 30 minutes, which caused the load to reach the black after they filled up the last VPS sposts on katarn. It still puzzles me as to why the server could not handle that. Especially since they are crons installed by the CP.
We located those 2 days ago and saw an improvement until this afternoon.
What makes it so bad is that I am the one who has to notice when the server or the node is offline, I can't image what happens while I sleep.

In fact, there have been countless mornings where I would wake up and have several messages from my clients saying they where not getting any email due to the server being down.
You have no idea how stressful that is knowing my clients are not happy.

I love the fact that you get good attention from Greg, but when it matters the most, where is Tech support? When things are the worst, they take the longest to respond to the tickets, to even restart the server when my VZPP was not responding.

Yes, I'm frustrated.
But the only thing that would keep me here is to somehow get a second VPS in another location (Houston + Atlanta) to create a semi-balanced load server.
But I don't even know how feasable that is, I've never heard of anyone doing that with a VPS, only dedicated.

[rainboy]

Hi,

I can understand its frustrating... just wanted the idea out of peoples head that a DOS is something easy to handle if you have a server in one datacenter.

I also understand that you have a lot more problems as just this DOS hitting you now, which makes my understanding of your point of view a bit clearer. And i am sad to hear you have so many issues with the CP.

May i ask which VPS package you have, and if you are not running out of memory at the moment ? or are you talking about Cpanels cron jobs on all the VPS-systems on the hardware node together creating this issue ?

Kindest regards,
Patrick

[Highway of Life]

Memory has had enough space that I could quadrouple my sites and traffic and still be fine.
But it's the CPU usage that has topped out.
I have the Discovery plan.

I have to say, I'm sorry to have gone with VHCS. I do wish I had spent the extra money on cPanel. But we did not even realize VHCS was going to be a problem until we discovered those crons two days ago.
Hindsight is always 20/20

Though it does puzzle me as to why those crons where using up so much CPU power... it seems so odd.

[rainboy]

which crons are those ? which scripts, cpanel runs different cron's i.e. for backups but also for updates, and even worse, they start different scripts from such a cron file, do you get output of these tasks running, and did you check if there are any errors in them. (i.e. recompiling some updated program which keeps looping ?)

Anyways Discover has just the 'requirement' for Cpanel/WHM i decided to sit one step up to avoid memory issues.

Then till the start of this week, i did have problems with Apache, Mysql and php (or a combination of it) by leaking memory, causing my website to go above the bought memory. I decided to upgrade to php under cgi for completely different reasons, and i saw this problem was gone as well.

Kindest regards,
Patrick