iBurst very slow

Noticed today (1st Jan) that via Northcliff tower, the interrupts - ie: b/w stops but signal stays strong - is continuing - just enough to disrupt the more finickity of downloads, which wont tolerate a five second or longer 'pause' in reception.
Switched to Rosebank tower, and thus far only had one brief interrupt, which wasnt enough to mess with the http download..

Also noticed, and perhaps other users can confirm/deny - there seemed to be a clear connection between the start of different protocols being used and the sudden 'drop' in signal-bandwidth.. ie: I'm downloading solidly via http, and I open an FTP - and theres a clear stutter in the signal, as if some fast calculation is going on in the mind of the transmission/router/base-station.
This would occur in various combo's of thread 'A' being downloaded, and then a second thread 'B' starting - usually ftp, but also in combo with an IRC 'get' - and also in combo with a flashget 'get'.
I'd clearly see a total break in data in/out with the start of a second protocol.
(Thats my take on it..)

Via Northcliff it does appear that there's still a problem - assuming that a phenomenon which 'interrupts' downloading, is seen by IBurst as a 'problem' :)

PS: Their Orcs arent too bad in terms of helpfulness, by the way - altho I'm not so sure about the IT-free Orc known as Portia.(And having a checklist of 'things to ask the customer when there's a problem' - isnt enough to compensate for a lack of IT knowledge.)

PPS: And Luis is a decent bloke who needs to gather up some equally tech-savvy and customer-friendly techies around him - as when things start going wrong, its not good enough to have an admittedly helpful support staff who can only say "I'm looking at the monitors now, and everything looks fine" when its clearly not.
They need a couple more techies lurking around, who have skills to run toolz if/when customers whine. Help staff who can only look at monitors and say "I'll have to phone Luis" or "When Luis comes in I'll tell him" means there's an awful lot riding on just one guys knowledge and ability to be around..

But looping back: to Iburst users noticing what seems to be a 'sluggishness' or 'unresponsiveness' - see what kind of protocols you're doing or starting, and watch your send/get data transmissions at the same time - to see if you also get a direct drop off at certain actions.

I find setting a permanent ping going, or starting a huge local download which takes up half the available b/w, seems to override the drop-off - allowing me to sneak in the download without a drop-off problem - but its a brutal method which is irritating as hell. Shouldnt have to do this. So its an ongoing error, which they're not seeing..

(While writing this, its just happened twice via Rosebank tower - but each time luckily, the http downloads I was pulling, decided to continue..)
 
Last edited:
Also experiencing the same crap. I'm using the Bedfordview/Primrose tower and it started yesterday morning. How can I switch towers?

All I can say is it sux.
 
Not sure of the geography where you are, or the tower(s) you have available - Im lucky enough have a big brick block of flats, on one side of which I can get rosebank, and from the other side, and a loooong ethernet cable, it gets northcliff..

Ping reply on IRC during the last 'interrupt' = "XXXXXX ([email protected]) had a ping reply of 29.672 Seconds!"
Ping reply on IRC when the signal came back = XXXXXX ([email protected]) had a ping reply of 3.547 Seconds!"
edit: Ping reply NOW, when signal seems stable and fine again -
"XXXXXX ([email protected]) had a ping reply of 1.641 Seconds!"

rebel - I'd suggest that phoning the help desk and telling them that you're having the same problem, is vital. So that they (and Luis) see that its not just one whiney customer, but a genuine Network problem being experienced by other users as well.

Thus far they seem way more willing to interact and respond to customers, especially on tech-issues, than Sentech, so make that phone call - coz otherwise they arent going to know there's a real problem..

To override the drop off, or at least 'punish' IBurst while I sit with no way to download anything, and broken http downloads, I use a browser link to one or another 1gig file resting at saix's game site..
 
Last edited:
LG, assuming that the problem you are experiencing is the same as the problem I have been getting (not willing to say it has disappeared yet), I can definitely say that it happened when the only thing I was doing is browsing mostly locally (no downloads going except a/v update checks every 3hours).

Not to say that I have done 0 downloading at all since 2004-12-23, but it has definitely struck me when I have only been browsing.
 
ic - hmm, maybe I'm linking it to various actions/protocols, because Im always fiddling and grabbing things when the signal drops - still, if the help desk gets at least 3 or 4 people wanting to speak to Luis, and having a roughly similar story, from different suburbs, this'll prod him/them into seeing that its something needing to be addressed.. (although earlier when I spoke to the Orcs, Luis' phone was off..dunno if he'd be coming in today/before monday)
 
Help yourself, help us all: agree to test iBurst by emailing [email protected]

LG, I agree - too much seems to rest with poor Luis (Anakin), they need to accelerate the cloning of Luis process.

As a constructive suggestion to all of us iBursters - if you have not yet decided to Quality of Service test iBurst, please get off that apathy fence you are sitting on, and email RPM now at: [email protected], the more people involved the more undisputable the results will be (whatever they will be) :)
 
Believe or not, there are other Luis at WBS. Your Luis was struggling somewhat with his senses today. :cool:

I checked Bryanston and Northcliff for IC and LG moments ago. Bryanston seems to be humming along nicely, with lots of users getting their share of speeds.

Northcliff appears to be experiencing some kind of interference, although users on it seem to be humming along nicely with speeds too, but some packet loss could be experienced. I will notify the radio guys to check out the interference - I hope they have recovered their senses today.

EDIT: I spoke to another Luis, aka Sugar, and he was aware of Northcliff and had been busy on it today. He hopes to have issue resolved by Sunday morning.
 
Last edited:
Thanks Luis aka Anakin, so Northcliff people should ask for Luis aka Sugar when calling between now & tomorrow...

This cloning programme is confusing me :)
 
Ok, to avoid confusion - by other Luis, I meant clones of Luis. There is only one Luis at WBS, but there are clones of him at WBS, e.g. Sugar and rest of technical team.

Normally, clones and jedis do not liaisse with iBurst subscribers directly - Sihen and his Customer Care team manage this aspect. Occasionally, clones are asked to assist as well.
 
Last edited:
I would have thought wireless coms hardware developers would try rather harder to make their designs less sensitive to interference :eek:
 
Hmm, seeing as the Star Wars analogies are flying here, with clones and Jedi's - can I be the first to suggest that maybe the Help Centre phone answerers at IBurst be called Ewoks or Wookies.?
(It has a certain ring to it, for instance "According the portia wookie, the towers seem fine right now").. although I miss the Orc analogies :P

Sunday morning, 8.30am, the 'stuttering' of bandwidth is continuing, messing up downloads. Ive had to re-download a 36meg file four times sofar.

EDIT: As I'd noticed previously, when running pings, this seemed to stop the connection dropping for some reason - but who wants to have a dos box continuously pinging, in order to have a stable and decent bit of useable b/w - have found that having a continuous http download going, also forces the issue and stops the stuttering, which interrupts IRC d/l's.
So a continual http d/l - whether local or overseas - seems to force a solid connection and allow other apps (specifically IRC) access to a stable b/w. Not a good 'fix' - but it works for me..its a b/w guzzler, but its that or nothing..
 
Last edited:
Hmmm, I took my UTD to a friend's place last night in Randpark Ridge, I assume it connected to Northcliff, there were no problems at all, which surprised me.

LG, you are only ever using RJ45 as opposed to USB with your UTD right? (I assume due to your loooong cat5 cable vs max 5m USB cable). I know this might sound crazy, but try USB when it happens again on the side of your flat closest to your pc (an experiment to confirm or deny another theory).

I think I prefer the Star Wars analogy applied to the WBS Helpdesk, they haven't reached the levels of disservice inflicted by Sentech on its customers.
 
Last edited:
Ok, Luis (aka Anakin) bad news, the Unresponsiveness problem has just struck again (btw I am back on Bryanston basestation not Northcliff).
 
Just confirming that the service is extremely slow today and yesterday as well. Hopefully it'll improve soon...
 
ic - I get the state that I see as 'stuttering' of the bandwidth, and which perhaps is the same or similar to what you see as 'unresponsiveness' - regardless of whether its ethernet or USB. Tried every variation possible, short of killing a chicken and conducting voodoo experiments over the modem..
 
If that fails, try santa-ria, apparently it is closer to the dark bandwidth spirits than old pleb voodoo.

No seriously, last night I used USB in RR - preumably Northcliff tower, connection was used all thru the night until about 5am, no stuttering/unresponsiveness. If you have tried USB then it should rule out the theory that MTU changes will make any sort of difference.

I am considering a packet capture with EtherEal/EtheReal (WinPCap) the next time it happens.
 
The attack of the stuttering/unresponsiveness again- decided to initiate a manual linespeed test when it struck- download speed shows as 18Kbits/s and upload speed 0 (results timestamp 03 Jan 2005 13:21).
 
After puzzling a Wookie at the helpdesk, I spoke with a certain Jedi at Iburst, who said they couldnt do anything about what I was experiencing via MIRC - I mentioned this thread and that others are experiencing variations on what Im getting, but via http - and the Jedi effectively shrugged and said they'd look into it.

Mention was made of the fact that IRC (DCC) is a dodgy protocol and dependent on the other persons pc and speed - but still - there SHOULD be a solid pool of bandwidth available that doesnt stutter or hesitate regardless of what one might be doing with it.

Methinks a followup call to Iburst might be in order ic - especially if you have toolz which can/have recorded the phenomenon..

Currently I'm using a ton of their b/w on a perm local d/l to stabilize my 'gets' off IRC, Ive mentioned this to them, and the fact that I dont like doing this - but they dont seem to get the wastefulness or the horror I have of wasting so much b/w, (perm d/l @15-25k local - in order to get x3 7-8k steady IRC d/l !) but Ive told them, so its not my problem.
I was just told to the effect that 'once capping starts you wont be able to do that' - which perhaps missed the basic point - namely that something is wrong and is requiring a solid download to stabilize signal. Its either that - or no IRC downloading at all..

Once capping starts however, it'll be interesting to see whats left of the Internet for IBurst users to access in a reliable fashion - if p2p is blocked, and MIRC stays a stop-start-erroring-out experience.. then it would pretty much be just http and ftp left.
Even though it may be 'fast' - there's precious little these days on web and ftp, to justify the pricetag..
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X