10KB/sec is not even close to "100Mb guaranteed"
After almost 72 hours an intervention for "Server check" was scheduled.
An hour later the following message was appended to the ticket: "The replacement of your network cable and switch port has been programmed." This sounded promising, but it also stated "This intervention was planned as a result of the errors detected in your MRTG graphs." Whether this implies that it would have happened in spite of my ticket or not, I don't know.
Another hour later and a warning that the server would go down in 15 minutes for the "Server check" intervention arrived.
A little over an hour later still, a message arrived noting that the intervention had been completed. It contained the following message: "Here are the details of this operation: Motherboard replacement".
I failed at trying to make sense of any of this. I don't know why it took the technicians so long to respond, whether they intervened because I raised the ticket or because they spotted "errors" in the MRTG graphs (which I couldn't see), whether they replaced the network cable or switch port at all, and found that it didn't help, or just decided to go directly for a motherboard replacement instead. What I do know is that whatever did actually happen did, in fact, fix the problem.
Now I have full speed in both upstream and downstream directions, at last, though it has been almost two weeks since I purchased the server. It's disappointing that the communication was so poor and that it took so long to address the problem but at least whoever finally looked at it knew what they were doing and made things right.
Originally Posted by
Marks
Kimsufi server has this network availability compensation when it come to case of Network problems. That covers problems with the network.
If this isn't a network problem what kind of problem do you think this is?
After 24 hours nobody has so much as bothered to even acknowledge that any ticket exists.
DedicatedPros
14-10-2009, 16:25
Yes, please post as much relevant information, otherwise we can't really help and you just end up trolling.
BTW, the engineers will look into your ticket as soon as possible. Then we'll have more information about what's the problem.
When inquiring about compensation for the unavailability rendered by the network failure he cited a 10% chance of succeeding in such a claim given that the box is just a Kimsufi, which they don't really care about because they have pro servers to deal with.
Kimsufi server has this network availability compensation when it come to case of Network problems. That covers problems with the network.
Generally speaking, it's not possible to compensate for problems in Kimsufi, that's in our guarantees. That's the compromise you get into in getting a Kimsufi server: cheap but it doesn't come with proper intervention times, guarantees and compensations as a Pro servers do. This is probably the biggest difference between Kimsufi.co.uk and ovh.co.uk.
I've seen cases in the past that, for very extreme cases we've given certain compensation, but it's not in our conditions. but that's not even 10% of the cases, these extreme cases happen less often.
Post a 100mb ish test file we can download from your server, we'll be able to give results from different locations.
Now, I'm not sure if this asking for help or making a complaint? If it's asking for help, it's akin to saying "My car only does 3 MPG" without giving any further details. Could well be you're driving on flat tyres, no?
Anyway, what's to OS? And what does a tcpdump give you? (A connection diagnostic's best friend, IMO).
Definitely sounds like a NIC issue or some form of blocking being done by the network.
It has been confirmed by people all over Europe. Yesterday the best they could pull was about 10k/sec. This morning it's generally zero. Occasionally some data squeezes through. My SSH sessions freeze any time there's more than a single screens worth of data to display.
Although the support guy I spoke to on the phone agreed to expedite the ticket, he hasn't actually touched it.
tracert's
pings
screenshots
They all help
Would be nice if you could actually post some information showing the issue instead of just complaining away..
Tried updating the NIC drivers?
Have you tried a connection from multiple places to see if the problem exists for everyone?
Keep bugging support until they do something about it.
Thanks for assuming that I'm too stupid to comprehend a telephone conversation, however what he actually said was that in 90% of cases like this I would not be successful in making a claim for compensation because it is just a Kimsufi. I did some basic math and transformed that figure into a 10% chance of success. Therefore it is impossible that I misheard him to have said what you claim.
gregoryfenton
14-10-2009, 09:51
Originally Posted by
Bilge
When inquiring about compensation for the unavailability rendered by the network failure he cited a 10% chance of succeeding in such a claim given that the box is just a Kimsufi, which they don't really care about because they have pro servers to deal with.
I wasn't there, or privy to the conversation but I suspect you have selective hearing or reading.
From the kimsufi.co.uk homepage:
Maximal refund per month
(% of the server price) 10 %
That sounds like what you were told, not that you had a 10% chance of getting a refund because the staff aren't interested.
After calling support the technician noted the packet loss while pinging my box. He surmised that either my processes are running to 100% load on the CPU or that my box is under attack. After explaining twice that my load is a solid 0%, he noted spikes of incoming traffic on my MRTG graph and decided instead that that indicated an attack. After explaining that that was just me doing speed tests he told me that it was still going to be one of those two things.
When inquiring about compensation for the unavailability rendered by the network failure he cited a 10% chance of succeeding in such a claim given that the box is just a Kimsufi, which they don't really care about because they have pro servers to deal with.
This morning the problem is so severe that I cannot use SSH any more, nor download a web page over HTTP or file from the server over SFTP. Even with the program FlashGet which can attempt to download a file over HTTP with 10 threads, after reaching an all-time high of 30k/sec, only a few seconds later it trips over itself and completely dies bringing the transfer to a grinding halt.
Incidentally, downstream tests show that the NIC is capable of a full 10MB/sec, so only the upstream is affected.
DedicatedPros
14-10-2009, 07:51
Blame others without even thinking to contact support, because something may or may not be causing the slowness.. that's very nice of you.
He's trolling. Take a look at his previous posts
I'd recommend submitting a ticket so a technican can check your network card. I have always had excellent speeds with OVH.
I've had the most difficult time these past few days setting up my new box. I am migrating from my old one. Now that I can perform some speed tests it turns out that it is because I can only pull about 10KB/sec, inconsistently, from the network. I'm also getting packet loss, enough to cause SSH freezes at times. Independant tests from other users via HTTP confirm the same readings.
I imagine it is because I am sharing the network with a bunch of "torrenters" and "seedboxes" as I have heard this company's main business is with torrent warezboys.
I can't even run `top` properly without it freezing or at least "lagging" at any time during the day. I cannot begin to describe how infuriating this is. At least my old box never had network issues to speak of. Now I'm regretting the migration altogether.
