Portal Home > Knowledgebase > Articles Database > Continuum/packet loss
Posted by proclanh, 02-09-2011, 11:07 PM Anybody having issues here with packet loss and dropped connections. This has been going on for a week now and seems to getting worse. We were told faulty core networking hardware is the problem. You would think a new DC would be able to fix this issue in less time. I can't even transfer a file from there to colocrossing without the transfer timing out.
Posted by gcphost, 02-11-2011, 11:56 AM I DO! Massive packet loss during busy times, not so massive - but still there any other time. Downloads are POOR - 22K per sec, my fios can upload faster then that! I asked them EVERY DAY if this will be fixed I know its not all their servers, I talk with 2 other co-located customers there who have no issues, anymore, they had issues for a day or two - then were fixed. Continuum will be loosing my business if this isnt fixed by the end of today.
Posted by scooby2, 02-11-2011, 10:47 PM I've been leery about moving my services over. I've had my server colo'd from about a month now and there has been one couple hour outage and numerous bouts with packet loss. I did not notice if there were problems today but on 2/5/2011 I had major packet loss for hours. Continuum never did respond what the problem was. They just said "we are looking into it" and closed the ticket. Time to shop around I think.
Posted by proclanh, 02-12-2011, 03:25 PM Thats what I get from them, also this has been going on for a week now and they respond with there working on it and close my tickets. Whether they oversold there network or it's a core router issue telling me it could be a week to fix with no ETA is not a very good answer coming from a so called "state of the art DC". I'm about ready to drive there and pull my servers.
Posted by gcphost, 02-12-2011, 06:58 PM Their site says: "Our SLA guarantees a low-ping, low-latency network at all times." Tho my SLA does not guarantee this at all. They told me, today, that on Monday they will have final date for the fix - they are aiming for the 19th but that may change. So pretty much another week of shotty service is a guarantee - a fix - not guaranteed. If they cant fix a hardware/network issue within a few days, let alone WEEKS then I have no trust that they can fix a MAJOR issue such as a fire or some large hardware failure. Looks like myself and a few other people will be pulling out very soon.
Posted by proclanh, 02-12-2011, 09:11 PM There 50.00/month colo special on here does not have an SLA agreement but they won't tell you that up front. In order to be bound to there SLA you will need to sign a 3,6 or 12 month contract/lease with them. If there telling you the 19th I will be pulling my servers within a day or two this is BS. Sounds like to me they oversold there network capabillities. We get packet loss all day and dailup speeds on ftp transfers.
Posted by timburke, 02-12-2011, 09:20 PM Everything seems fine here. It could be that I'm contracted with them and therefore have an SLA...
Posted by proclanh, 02-12-2011, 09:39 PM So what your saying is if you have no SLA you get throttled down to a snail's pace? try transfering files out of there Data Center to another location and post your transfer speeds. We also use Colocrossing and I could not get the speeds to go above 100kbs and thats just across town.
Posted by gcphost, 02-12-2011, 10:14 PM I get 230.0KB/s to my backup server in Georgia. Since this started I typically only get 20KB/s during busy times, I used to be able to max my connection out. Not everyone there has the issue - I talk with other hosters who have no problems at all. My SLA only covers downtime - I have the $59/m deal that was advertised here. I think they're doing some false advertising. Here is the last reply I got: They cant even set a date to fix an issue that has existed for 3 weeks. Last edited by gcphost; 02-12-2011 at 10:22 PM.
Posted by polycomvoip, 02-12-2011, 11:14 PM What kind of SLA do you want for a service your paying 50.00 a month for?
Posted by scooby2, 02-12-2011, 11:16 PM No packet loss would be nice. As would an explanation about what is going on versus just closing tickets.
Posted by gcphost, 02-13-2011, 10:58 AM I'd want the SLA they advertise otherwise thats false advertisement - there is no asterisk here, this is cross the board advertised to everyone.
Posted by Packhead, 02-14-2011, 01:21 AM It's been a small irritant to me, but I know they're working on it. That being said, it seems to mainly impact TCP connections on my server. I guess I'm not feeling the hurt as much since most of my traffic is UDP based -- all of that seems to be still functioning just fine. Also, if you read the actual SLA, I don't see anything in there about guarenteed rates/latency/etc.. I'd like to see it, but all the SLA provides is guidelines for "failure that results in the loss of Services to Customer". Maybe I'm missing it. I agree the site says stuff about their low latency SLA, etc... But I don't see it in writing. And if you are angry at them for low latency and saying it's against their SLA, and the SLA doesn't even indicate it, then I'm not sure you have much ground here. Either way, I'm still incredibly happy with Continuum. I'm going from month-to-month to a 3 or 6 month term, and bringing two more pieces of equipment in later this month. I'm very pleased with their responsiveness to tickets and willingness to work with their customers. You show me a datacenter in Chicago for a better value and I will be surprised.
Posted by gcphost, 02-14-2011, 12:50 PM Same with my agreement with them. I'm just angry for the false advertising - I did sign up with them after reading their site and thinking we would have a SLA that not only covered downtime (standard) but also the latency issues. I see most people are probably happy there but if you're getting 20kbps downloads for 3+ weeks with no quick resolution, then maybe you wouldnt be so happy I'm trying to launch a new site and cant really do that with painfully slow downloads.
Posted by proclanh, 02-14-2011, 01:04 PM We didn't choose them for the pricing we put our main control panel server there to keep it out of the loop of our other 30+ servers we have located all over the world. We run game servers and this server monitors all of them including ftp access to the boxes. So since the latency issue clients can not access ftp or the control panel without it timing out. Trust me I tried to transfer a 400mb compressed file to the Dupont Fabros facility across town at different times of the day and my average speed was 100-120kbs. While I've been happy with the service since we've been there waiting two weeks to see a fix to this problem is not good for us.
Posted by pentiumone133, 02-14-2011, 01:52 PM Just posting to say that I dont seem to be having any issue at all. I'm also on the $50/month colo special they ran a few months ago. Pulling a full 100mbit from FDC Denver and Chicago to a VM running on my esxi box in Continuum.
Posted by gcphost, 02-14-2011, 01:56 PM I've seen it on other servers there, they work just fine! Im doing 123.5KB/s now to Georgia - used to max my VPS out in Georgia but now I cant even get above 127KB/s So its like all of us in this 1 cage or rack with the issues.
Posted by pentiumone133, 02-14-2011, 01:58 PM What rack? It says in the control panel. me: Row C: C7 (Shared Rack) My file transfer just finished. Moved 8GB from Continuum to FDC denver at 11.8MB/s AVG
Posted by gcphost, 02-14-2011, 02:00 PM Zone & Cage: Suite 1: Default Cage Row & Rack: Row B: B5 (Shared Rack)
Posted by pentiumone133, 02-14-2011, 02:02 PM Maybe request they move you to a different rack?
Posted by gcphost, 02-14-2011, 08:37 PM Found out with our own debugging that its not the rack or cage - but a router, same cage/rack - 2 servers, 1 has no issue, 1 does, difference is 1 router. Sad it took some friends 5mins to debug their problem thats lasted weeks.
Posted by Packhead, 02-14-2011, 08:58 PM They know what the issue is and they've ordered stuff to replace it. Granted I do agree it should be in place quicker than it is, but maybe that's the only maintenance window they have. Anyways, looking forward to having it fixed up.
Posted by gcphost, 02-14-2011, 09:04 PM I think a month to fix a router is to long - Im pulling out, just sucks I can only transfer at 120kbps - its going to take a couple days to move my vps's over. ** as I typed this I got a reply to a ticket that said they OVER SOLD so I'll make it easy for them to get more bandwidth sooner then their lame deadline of the 19th.
Posted by Packhead, 02-14-2011, 09:06 PM /me shrugs. Good luck.
Posted by gcphost, 02-15-2011, 01:19 PM I should have bragged about the 120+ kbps, today its 30! Friend in the SAME RACK getting 11MB/s.
Posted by fraghost, 02-15-2011, 09:31 PM I do not have these issues. I DO have an sla though. 2011-02-15 19:29:34-- http://67.159.44.209/100MBtest.zipnnecting to 67.159.44.209:80... connected. TP request sent, awaiting response... 200 OK ngth: 104857600 (100M) [application/zip] ving to: `100MBtest.zip' 0%[======================================>] 104,857,600 11.0M/s in 9.0s 11-02-15 19:29:43 (11.1 MB/s) - `100MBtest.zip' saved [104857600/104857600]
Posted by gcphost, 02-15-2011, 10:08 PM I can download at 11MB/s no problem, upload - thats different. Test me vs you - to my fios connection. [root@new-host-2 cron]# wget http://gamecp.com/100MBtest.zip --2011-02-15 06:00:48-- http://gamecp.com/100MBtest.zip Resolving gamecp.com... 74.121.182.84 Connecting to gamecp.com|74.121.182.84|:80... connected. HTTP request sent, awaiting response... 200 OK Length: 104857600 (100M) [application/zip] Saving to: `100MBtest.zip.1' 16% [==================> ] 17,692,861 224K/s eta 6m 41s [root@new-host-2 cron]# wget http://67.159.44.209/100MBtest.zip --2011-02-15 06:02:13-- http://67.159.44.209/100MBtest.zip Connecting to 67.159.44.209:80... connected. HTTP request sent, awaiting response... 200 OK Length: 104857600 (100M) [application/zip] Saving to: `100MBtest.zip.2' 63% [=======================================================================> ] 66,183,485 2.20M/s eta 17s me - 203K/s you - 2.2M/s I asked if they could change my port - or whatever - they said - no - it wont matter Feel free to dl my test and see what you get - I have plenty of BW this month, lol.
Posted by fraghost, 02-15-2011, 10:56 PM hmmmm, I am getting about 500K/s to me. I do have an SLA, but I don't think there is anything in there in reguards to throughput. I am not dropping any packets. My critical sites are hosted with Ubiquity. Kind of annoying, but I'm not paying that much either. I may consider moving, but we'll see. Good thing Chicago has a lot of alternatives.
Posted by FredFlintston, 02-16-2011, 01:41 AM I've had a cab with them almost since they opened, and I'm in the same boat as you guys. Off-peak bursts to 1.3M-700k, peak bursts if you're lucky up to 300k. Both TCP & UDP are affected. What really bothers me is that they're totally brushing it off, "The resolution is tentatively set for 2/19/2011." Tentatively? You mean you don't actually know what's causing the problem, so you're throwing parts at it. Way to really inspire confidence! If they weren't so damn cheap, my equipment would have been gone last week. But if this persists past the weekend, I'm gone.
Posted by trancemode, 02-16-2011, 08:31 AM So is this affecting ALL cabinets or racks? Or just certain one? I don't think I have an issue, well hopefully not. going to give it a check. Wasn't aware of the network issue as well.
Posted by gcphost, 02-16-2011, 09:49 AM We proved its not everyone, we found 2 routers - 1 router with the issue - 1 without. We've had to basically guess at the issues as Contniuum hasnt said much about it. Contniuum was not very up front about this issue - they would never tell you it even existed.
Posted by fraghost, 02-16-2011, 11:12 AM Omitting network issues, I'm quite upset about that. If they would have been upfront and informed me that there we network issues, I would have stayed, but I have canceled services. I wish everyone the best of luck, and if you decide to move services away, let me know which DC you moved to.
Posted by gcphost, 02-16-2011, 11:42 AM Direct from them is as reliable as it gets I'll share more
Posted by fraghost, 02-16-2011, 02:22 PM Agh, I gotta make a decision. To leave and move elsewhere, or stay and wait for a now firm fix on the 19th. Hmmm. I setup a cancellation, but Shauna is just so damn nice! I wanna stay, but I don't want to deal with these issues, but if this is a for sure fix, you simply can't get a better deal elsewhere. Decisions decisions. Anybody else going to wait it out or move?
Posted by gcphost, 02-16-2011, 02:35 PM You trust them?
Posted by fraghost, 02-16-2011, 03:55 PM I think that I will wait until Monday to make that decision. What about you? Do you trust them? Are you staying? I don't know how I feel as far as trust goes.
Posted by gcphost, 02-16-2011, 05:33 PM Really hard to want to stay, I've had to invest 2x more then the measly bill I pay them to keep my services alive, and stable. They've cost me to much as it is now - I'd be scared of these issues happening in the future and never being told - as is the case now. To get me to stay would probably cost them more then they want to pay.
Posted by fraghost, 02-16-2011, 07:23 PM I can certainly understand your logic. What alternatives where you looking into? How much space do you have with them now?
Posted by proclanh, 02-16-2011, 08:45 PM I already put in a cancellation. I was probably the first to notice the issue because our clients log into ftp to admin there servers it's a little different than a web site where you really don't notice packet loss.
Posted by gcphost, 02-18-2011, 09:57 AM Where are you moving to? ** im ready to move today! if there was another co-lo facility in Chicago that I havent dealt with (steadfast, gowebman or coresite) then I'd move right this second!
Posted by fraghost, 02-18-2011, 11:01 AM Have you tried ubiquity?
Posted by gcphost, 02-18-2011, 11:08 AM Ya I looked today after a good recommendation but their to costly, their affordable plans are all for 1mbps, I eat 3mbps constantly so 1mbps wont cut it - to get 5 is like $400+ a month, a bit to expensive.
Posted by fraghost, 02-18-2011, 11:14 AM I saw a special they were running on WHT a few months back.10u 10amps and 10 network drops for like $350 bucks per month. I think it was either 25mbps capped or 5tb/mo. I wouldn't mind splitting costs with you if you were interested. =P
Posted by gcphost, 02-18-2011, 11:16 AM I would be - but I only need the 1u, seems @ 10 slots the cost would be fairly cheap? I'll ask around too.
Posted by gcphost, 02-19-2011, 11:32 AM ** i was wrong Last edited by gcphost; 02-19-2011 at 11:39 AM.
Posted by Celeborn1979, 02-19-2011, 03:52 PM After being with continuum for a year it seems they are turning into FDC. anyone else looking for a new provider? if so who are you using.
Posted by trancemode, 02-19-2011, 06:01 PM That really sucks! I am guessing i am on the other router. However their maintenance is tonight so hopefully its fixed
Posted by gcphost, 02-19-2011, 06:38 PM We're trying to move to Ubiquity - if you need space private message me and we will get you in the loop, looking at $75/m for 1u with similar offerings as what you get now - Ubiquity provides Continuum so we're talking good stuff but more overhead - so no overselling.
Posted by Packhead, 02-19-2011, 11:23 PM What? FDC is far FAR worse than anything I've ever dealt with at Continuum. I've rarely had downtime with them. And the kicker: whenever Continuum has been down I have always been able to talk to support or Tom instantly. None of the phone games like FDC plays, lack of responding to tickets, etc. FDC was a joke. Maybe it's me, but I'm not seeing nearly this level of "rage quit" worthy problems. Sure I'm seeing drops, but they're working on it. Sorry, things don't always happen overnight. I'm not leaving Continuum -- infact I'm sending them more stuff. I trust them. They're still growing.
Posted by Violent Injection, 02-20-2011, 02:06 AM Personally I am on the fence. While this is the first big issue there have been a few minor ones which is meh. But the thing that got me was when I asked about what was wrong a week ago they told me a router issue not that the network was oversold I personally don't like being lied to and that is what I feel they did.
Posted by Packhead, 02-20-2011, 02:49 AM Maybe I've missed it (I'm pretty tired), but where is the post that states they were oversold? If they were oversold everybody would be having the problem (assuming it's their AboveNet link), not just certain people behind certain routers. They said they were using this outage to upgrade their links (probably to avoid a downtime in the future for it) as well as replace hardware. I think there's some jumping to conclusions here without a solid answer from Tom (which I know he will be provide since they aren't some shady fly by night datacenter, like other downtown Chicago DCs seem to be). I honestly believe they had equipment issues that they are resolving, and at the same time decided to just go with the link upgrades. Again, if I missed the post from continuum stating they oversold the links, my apologies. However, in all my discussions with them, the issue has been strictly hardware on their side of the fence. I still couldn't be more happy with them. They provide a great service at an incredible value. Try and get the prices you do at other colo facilities in downtown with the uptime we've had -- it just doesn't happen. The power issue a few weeks ago and this have been the only trouble I've had with them in over 6 months. That's pretty friggen good, especially for what I'm paying. Edit: I also appear to be back online.
Posted by gcphost, 02-20-2011, 03:03 AM I did in my comment. They have said all along its a hardware issue, waiting for hardware to be purchased - etc. How long should it take a proper datacenter to fix a faulty router or switch? 24 hours? 1 week, 3? This took at least 2 weeks from when I notified them - that was already 2 weeks into the issue, I've noticed this issue for 4 weeks (from client reports). Did they tell you they were having problems or did you figure it out on your own like most of us here? I agree they have been good - only issue in a year has been the recent downtime, their great on most of my support requests and even replaced some hardware for me, never had any issues till now but this is quite major 3 weeks of 30kbps downloads hard to deal with.
Posted by scooby2, 02-20-2011, 03:22 AM Looks like they just finished up at about 1:10am Central. Things are much better right now and hopefully they stay this way. Overall I am satisfied with Continuum however I do wish they would have been more forthcoming about the cause of the packet loss problem. I still have not seen an email or ticket response stating what exactly the problem is/was. After pushing more they finally did tell me they would be replacing a piece of equipment on the 19th.
Posted by fraghost, 02-20-2011, 07:31 PM "range quitting" is an exaggeration. I agree Continuum is a lot better than FDC. It's a business decision to leave, has nothing to do with quitting because one is "angry." In fact, my cancellation notice was very professional. I will not be renewing with Continuum. A suggestion was made for people that are also making the business decision to leave, check out Ubiquity. May cost a little more, but I think they're a little more prepared as far as available power and bandwidth... Just an opinion, nothing more.
Posted by tchaffin, 02-21-2011, 05:41 PM This issue has been fully resolved under our 2/19/2011 maintenance period. We do apologize for the inconvenience and appreciate your patience throughout the matter. If you have any additional questions or concerns please contact our support department directly and we will be happy to address your needs. -Tom Last edited by tchaffin; 02-21-2011 at 05:41 PM. Reason: typo
Posted by gcphost, 02-21-2011, 06:35 PM Can you clarify why it took over 3 weeks to resolve this issue Tom?
Posted by tchaffin, 02-21-2011, 06:46 PM It took almost exactly 2 weeks to resolve. We required a maintenance window that was 1) convenient for our entire customer base and 2) allowed us timing to prepare for a smooth maintenance window. Unfortunately, maintenance windows will always be inconvenient for someone. -Tom
Posted by FusionNET, 02-21-2011, 08:43 PM Tom has always pleased me in my needs, and provided a great service. I do understand the fusteration with some people having the outage, and problem they where facing, but I do not believe it's because they where oversold, and i have never seen anything official coming from them or the providers saying they are oversold, and oversold is quite a scary term, and to each person there is a different meaning, if the guy who stated that, can please provide justification on that please clarify or where you think, or were you saw that they where oversold. I would just like to see this. I was with them for almost a year and 3 months, and they where hosting everything within the FusionNET origination for the longest time, as our main provider and still hosting some things within the FusionNET origination, but not under my needs anymore its handled by a different member of the team. But they where always very helpful, and reliable, and had a great service provided to us. I would recommend to anyone. I must do say, we never had a major problem with them like we've had with some other providers, and for at least the time I was personally the client with them we had a amazing service. I do agree with the business key about, whats better for your business is what you need to do and if they are not suiting your needs it's best to transfer, the same thing we did, well personally I did, I moved to another provider to fit my needs better, but when I was with them personally, they where everything but anything related to FDC. As I can't say much about FDC, because I was only a client with them for 30days back in 2010 around January, and left the month after due to connectivity issues, but FDC is no where near the company Tom's is. I believe they provide a top-notch company, and I recommend them for anyone, like I always say with hosting company's, You can't please everyone's requirements and needs, Not every host is going to be the perfect host for everyone, Always check to make sure they fit your requirements 100.10%, and there falls make sure they don't bother you as much. Keep up the good work, Glad to see everything is back stable. How's business for the new year going Tom? It's been awhile since we chatted Note: I personally don't see nothing wrong with FDC anymore, as they seem to fixed some of there issues, but I really cannot comment on that as I'm no longer a client with them and wasn't a client for very long, as my service I did have with them, For the 31 days I had the service, it was online for 22, and nothing on my end was reported back to me was my fault. But from friends, and reviews it seems they are coming back and doing okay for the new year? Good luck to clients, and Tom, I hope everything stays great as it always has been for us! Be a pleasure to speak to you again Tom, Thanks, Shane.
Posted by scooby2, 02-21-2011, 11:10 PM Thanks for the update Tom. Things have been much better since the maintenance was completed.
Posted by gcphost, 02-21-2011, 11:31 PM Ive been doing 90mbps for several hours, seems back to normal here as well.
Posted by Violent Injection, 03-06-2011, 04:51 PM Seems to be having issues again.
Posted by Packhead, 03-06-2011, 05:18 PM I've had a few customers complaining since the last window that there's been higher latency on their TS servers, and I've had a number of times where I've been just down for a few minutes. Seem to have at least one very small outage a day now. Not sure what the deal is. I just had a support ticket open and got a response back that the outage about 20 minutes ago was due to a DDoS attack on one of their clients. Seems one of their clients keeps getting DDoS'd... time to tell that client to stop getting people angry at them.
Posted by Violent Injection, 03-06-2011, 05:29 PM Or time to get rid of that customer before everyone else leaves. I am getting close to that point.
Posted by tchaffin, 03-06-2011, 05:31 PM The attacks are directed at game servers only. Unfortunately, this, ddos, seems to be how competitors, and those that get banned, resolve their differences these days. We are null routing the attacks as quickly as possible.
Posted by tchaffin, 03-06-2011, 05:33 PM This is directed at multiple customers. I'm fairly certain most of our game server customers have all received these from time to time.
Posted by Violent Injection, 03-06-2011, 05:39 PM Not sure, I have been running gameservers for 7 yrs never have been.
Posted by tchaffin, 03-06-2011, 06:37 PM This attack has been resolved and those that created tickets should have received responses.
Add to Favourites Print this Article