server.lu Throttling

[Six66Mike]

Has anyone else noticed throttled speeds on their servers? I confirmed it on 1 of my boxes, transfer 1200MB file to an OVH server never went past 4.2MB/s and dropped as low as 600KB/s. It took 18 minutes and 30 seconds to transfer.

Same server upload to another server.lu box got 7-9MB/s and was done in 2:39 so much better, reliable speed.

I've had other people show me their vnstat & MRTG, showing speeds never going above 3-4MB/s 24 hours a day for the last 3 days or so.

FAQ says "100 MBIT/s unmetered servers on average can transfer between 10-15TB per month outbound and 20-25TB inbound. The bandwidth allocation is not guaranteed/dedicated, and is supplied on a best-effort basis, which may result in some slowdowns during peak traffic times."

Is peak time 24 hours a day? There's no mention that intentional bandwidth shaping will be implemented but based on the tests and speeds, it seems to be the case.

Unrelated but the FAQ above that is about unavailability and knowing the facts. Several M servers have lost connectivity over the past few days, sometimes for a few hours & atleast 3 people reported faults. Still, nobody knows why it's happening, what the solution is or if one has been found yet. There isn't any disclosure or finding the truths that are passed on to the customer.

So is there throttling in place that nobody knows about or something else happening? Further up in the FAQ "...and similar are not covered by our unmetered services and that misuse will automatically incur a connectivity restriction after prior notification." If there is restrictions in place, nobody has been notified prior by any means.

[GalacticRhino]

I'm in the same boat here, and have spoken to a few others who have also noticed it today. Some of my servers haven't gone past 20 mbit (!!) in three days.

I have definitely had no contact about it - the thing that really pisses me off is that issues that have been ongoing for over a week (eg. a server which has been down) STILL haven't been fixed, yet I noticed over the weekend that MORE of my servers got throttled... great support there from server.lu, they have time to slow servers down to a near useless halt but don't have time to deal with urgent tickets.

It would appear all mentions of their "seedbox policy" (sharing 1 gbit between 4 servers) have been removed, as far as I can see. This isn't that anyway, this is a HARD limit at ALL times - there is no bursting, there is nothing. I definitely didn't receive prior notice, nor have a received a response about my questioning of it.

As far as I am concerned, server.lu are just shooting themselves in the foot once again - especially with this move. They call seedbox customers 'volatile', then they go and outright limit us when we are not even using as much bandwidth as the FAQ states is acceptable (between 10 and 15 TB, outbound - a lot of my servers didn't do those kinds of data, yet are throttled). I wonder what market they are appealing to now... their support is too poor to be considered by big businesses (seriously, if you are in charge of IT and you have a server down for over a week, with no sign of anything happening, how do you explain that to management?), their prices are too high for average enthusiasts and the amount of bandwidth they are now (seemingly) going to provide is too low for their original market of unmetered servers. This is definitely going to be an interesting discussion, and probably will become an interesting milestone on server.lu's history, for better or worse...

(I have signed up to post because emailing them has, once again, proven futile)

[Six66Mike]

I have a DS1000 that drops out under load, took about a week to get a reply and it said "It's fixed". That was the reply. No investigation, no details of the problem, the cause, the solution. So when it crashed within 24 hours I had no idea what the next step was :D

We have found out its a kernel panic so they said to fix our kernel, we've replied saying we're using their kernel, still waiting :|

But that's side tracking from my original point :) You got me riled up lol

After calling daily and emailing a few times as well I managed to get some other issues resolved, got a few more to sort out this week. When I have hit that "breaking point" the responses you get are good and do they work to fix it, but otherwise it's pulling teeth. Maybe it's because I logged so many tickets, I don't know :)

[eMoneyBags]

i have noticed it aswell. but luckily for me i purchased a box yesterday from a diff provider. so if they make this a habbit i'll be in the clear. :)

sucks for you guys tho :dabs:

[djfye]

I have the ds2000 and i noticed my upload never going above 3MB/s so I opened a ticket about it and got the response: Hello, Our system detected that you were using an average of 50Mbit/s, and your port was limited to 30Mbit/s.

[Six66Mike]

Well atleast you got a response I suppose lol. It would be nice if this was actually mentioned somewhere in writing to let people know, or at the very least an email sent when it happens. Pretty disgraceful.

/edit/: Apparently now it could be a bad config on new hardware if 1 person is right, or if another person is right the FAQ they had temporarily on their site saying heavy users will share a VLAN & 100mbit connection has happened. Still no clarification though, don't even know if its intentional or a bad config.

[v1r3d]

Same problem for me, my upload speed never gets about 600kb/s. :/
No replies to tickets.

[djfye]

well i got a message from them that said: Hello, We've checked the group where your machine was in, and we found out that the ports were not configured properly. Can you confirm a better speed now? Regards.
(lol they responed to my closed ticket-nice that they did though.)
since then ive noticed it go to about what it was before this all happened.

[Six66Mike]

Been testing mine all afternoon, doing the last few now and found a couple that are pretty bad. Gonna send in the server names to have them look to find out why. One is going around 400KB/s the other 1MB/s. Seems like some may be stuck at 5MB/s but others were 7-9MB/s which is closer to what we're expecting.

[kallejohan123]

just borrow this thread, hopes It´s OK for TS. cannot create own threads yet.

1; if I order a server @ Server.lu, do you think they can create it for two users witch each IP? Me and my friend wanna share.

2: which linux dist is best? debian? ubuntu?

3: Is still Server.lu worth the money?

we have about 80 euro every moth to spend on a seedbox - any other companies that it´s good for us?

sorry for bad english.

[manu1991]

^^ Id say stay away from server.lu , till they provide a reason for capping the above mentioned servers

If your country has OVH (uk , france , nederlands or portugal) buy from them , get a SuperPlan BestOF , for 60 Euro , you'll get 200mbps OVH to internet and 1.5 TB of space

[Six66Mike]

till they provide a reason for capping the above mentioned servers

I was hoping dwdm would have weighed in by now but he's been absent from the forums for a while. Still waiting for official response after yet another phone call on Friday :frusty:

[Ev0]

Looks like the server.lu bubble has popped

[xirvflux]

Looks like the server.lu bubble has popped

Seems so. dwdm has been posting here about seedboxes not being reliable customers... I think they must be from another planet if they think someone with their corporate website hosted there would tolerate tickets being unanswered or answered one week later, problems not fixed, invoices being sent early (they start billing the day they take your order, not the day their deliver, even if it's one week later), etc.

[Six66Mike]

By the look of it the timeline seems to kick in May 30th. Servers all moving sweet there, those with problems all seem to have slowed down from May 30.

xirvflux I brought up that last point and managed to get renewal days pushed back to the actual anniversary day they were delivered. In some cases it was 9-10 days, overall it was over 100 days difference combined.

Edit: *happy dance* well it's not resolved yet, but I got a response and some tests done on their end, and hopefully working toward a resolution. As long as the responses come promptly now and we can work toward a solution I'll be happy. If I'm waiting a week or so for a reply I'll be bummed out again. The responses I've receievd are good. Further cements my earlier comment that when you do get support, it's great. The problem is getting that support in the first place can be mind altering.

[eMoneyBags]

are you guys still having problems? mine seemed to only last for 1 day and has been fine ever since..

[Six66Mike]

Well they tested some of mine, pasted results of 48mbps on 3 servers, 29mbps on 1 and 96mbps on 1. The only one they said seemed to be off was the 29mbps due to "higher activity in this segment."

Apparently 48% bandwidth sustained for multiple days is acceptable and not warranting further comment. I also asked directly if this was intentional, network or configuration issue, and that has not been answered.

[FransTormer]

...I also asked directly if this was intentional, network or configuration issue, and that has not been answered.
Well, we do know from another topic that they were implementing throttling using VLAN's for servers that had speeds above xx% for more than yy hours on end. But judging by this topic, it seems quite unrealistic that every person here was using high bandwidth for longer sustained times and thus got throttled.

Hope speeds pick up for everyone and possibly server.lu will respond properly to tickets & questions.

[Six66Mike]

I directly asked if they are throttling and got this reply, which is word for word which I've seen other people get recently when asking the same questions.

"The servers in our Datacenter are wired in groups through 100Mbit/s shared connection. So if there are more customers who are consuming a larger bandwidth, others will remark that too with speed fluctuations.

If one uplink of a group gets saturated, the system ensures that every customer receives a fair share on of the bandwidth available for that group.

We can connect your server on a 'dedicated' line where you get more bandwidth, but you will have to pay for such an upgrade."

I've replied, waiting to find out how many servers share 100mbps ports & what this dedicated bandwidth is & how much it costs.

[Six66Mike]

Has anyone noticed even worse speeds after briefly having better speeds? I still haven't got a reply to the message I sent previously and the speeds have honestly gotten worse on a number of servers.

[lysine]

I've only got one server with them, but I still get good speeds on new torrents. here's one from a few minutes ago on scc:

http://i44.tinypic.com/2pqurld.png

[xirvflux]

Has anyone noticed even worse speeds after briefly having better speeds? I still haven't got a reply to the message I sent previously and the speeds have honestly gotten worse on a number of servers.

Right now I have two servers (happens to be in the same subnet so looks like a rack/router issue) down.

server.lu is becoming less and less reliable every day.

[Overlord]

Right now I have two servers (happens to be in the same subnet so looks like a rack/router issue) down.
Yeah my server is down, too.

[ClayDay]

All three of my servers are up and running great. I did notice some slower speeds last week, but that only lasted for about a day. Since then it's been great. On new torrents I'll max the upload out (11mb/s) and normally hit close to 10mb/s down.

[Six66Mike]

Right now I have two servers (happens to be in the same subnet so looks like a rack/router issue) down.
Yeah my server is down, too.

3 of mine are down, all K boxes by the look of it. Was 8 hours after they went down that I woke up & logged a ticket, but hopefully with multiple clients effected they put a bit more effort into resolving it.

[film42]

I have two K servers, one is fine, the other is unreachable.. it appears that even rescue console is down.

wtf lu?

[Six66Mike]

film42!

They stopped posting here when this thread was created :) They still post on WHT and claim to respond to hardware tickets promptly while ignoring ETA requests, but it seems even now hardware tickets are for the most part delayed up 1 week for a reply.

I'm hoping they notice an entire rack offline & haven't responded because they are fixing it.

Edit: I got some pretty promising answers again on Sunday, the boss was gone for a while and came back to a fun time by the sounds of it :D I've got my fingers crossed as the responses to my email were quite positive yet again. We'll see how it goes.

[darknforce]

Cheap but lack of support

[Six66Mike]

It's picking up though, still has its moments of frustration but with each passing week I start getting more & more confidence back.