Please Ring Eircom If...

Hank Scorpio

edit**

Andeh

Hey,
Ive been having the same troubles as the rest of ye. On the 3mb home package, and living in portlaoise.



Bloody disgrace.

Stephen

I'm getting the same sh1te in Kilkenny, but I'm on UTV... tis all Eircom bitstream though.

Marlow

Stephen wrote: »

I'm getting the same sh1te in Kilkenny, but I'm on UTV... tis all Eircom bitstream though.

Well, not quite.

Bitstream is one thing. It can be caused by Eircom, it can be caused by your provider.

To understand bitstream, take your example: UTV buys a DSL line wholesale of Eircom, Eircom hauls it back to UTV and UTV has to feed it with Internet of their own.

So your issues with speed can be happening at 3 places: Eircom's DSLAM (if it's fully loaded) and the line out of there, the fiber trunk that UTV bought of Eircom to haul the DSL lines into their network and then at UTV's internet connectivity. UTV specificly is not on the INEX, so also their traffic to the irish speedtest servers will always go down the path, of whichever carrier they bought their bandwidth of. Do a traceroute to for example "speedtest.airwire.ie" (Galway) or "westnet.ie" (Dublin) to see what route it takes (traceroute command on linux/mac/bsd, tracert command on windows).

/Martin

azzeretti

I have also noticed the same problem, right between peaks times. Bandwidth close to 500k and latency is horrible.

I am thinking of changing to BT or Vodafone but I am not convinced this would solve the problem, considering I would be connecting to the same exchange and the same line. How exactly does the contention work on the wire? Will the contention change if I move ISP even though the physical wire to the exchange is the same? What I mean is, is the contention ISP specific or line specific?

Assuming all this is contention that is, I have a feeling its not though.

Hank Scorpio

post up some traceroutes to rte.ie and boards.ie and it help to determine where the problem lies for you

Im think about switching ISPS too

Stephen

Marlow wrote: »

Well, not quite.

Bitstream is one thing. It can be caused by Eircom, it can be caused by your provider.

To understand bitstream, take your example: UTV buys a DSL line wholesale of Eircom, Eircom hauls it back to UTV and UTV has to feed it with Internet of their own.

So your issues with speed can be happening at 3 places: Eircom's DSLAM (if it's fully loaded) and the line out of there, the fiber trunk that UTV bought of Eircom to haul the DSL lines into their network and then at UTV's internet connectivity. UTV specificly is not on the INEX, so also their traffic to the irish speedtest servers will always go down the path, of whichever carrier they bought their bandwidth of. Do a traceroute to for example "speedtest.airwire.ie" (Galway) or "westnet.ie" (Dublin) to see what route it takes (traceroute command on linux/mac/bsd, tracert command on windows).

/Martin

Looks like its going through INEX to me... (note the awesome pings too)

Tracing route to westnet.ie [78.153.203.102]
over a maximum of 30 hops:

  1     1 ms    <1 ms    <1 ms  router.home [192.168.1.1]
  2   256 ms   261 ms   242 ms  194.46.193.65
  3   245 ms   322 ms   325 ms  193.95.145.26
  4   224 ms   232 ms   234 ms  vlan73.sw002.cwt.esat.net [193.95.130.217]
  5   245 ms   244 ms   249 ms  ge1-0.core001.cwt.esat.net [193.95.129.17]
  6   259 ms   252 ms   251 ms  pos2-0.br001.inex.esat.net [193.95.131.6]
  7   270 ms   276 ms   273 ms  ge-3-bdr1-dub01.39122.as [194.88.240.12]
  8   308 ms   316 ms   298 ms  ge-0-dr2-dub01.39122.as [81.17.240.70]
  9   306 ms   313 ms   303 ms  ge-3-dr1-dub02.39122.as [78.153.192.69]
 10   274 ms   275 ms   277 ms  192-100.colo.sta.blacknight.ie [78.153.192.100]

 11   267 ms   266 ms   275 ms  westnet.ie [78.153.203.102]

Marlow

Stephen wrote: »

Looks like its going through INEX to me... (note the awesome pings too)

It is indeed. After it goes through BT Ireland. Exactly what I said, it goes through whoever they are buying bandwidth of. In this case BT and westnet's box is in Blacknights racks (both connected to the INEX).

Looking at that, the contention could indeed be at the DSLAM.

/Martin

Stephen

Yeah, its always the first hop that has brutal pings in the evening.

Marlow

Stephen wrote: »

Yeah, its always the first hop that has brutal pings in the evening.

That would be the hop from the Eircom DSLAM to UTV. So the question then is, is it the DSLAM or UTV's trunk from Eircom (or inbetween).

The only way to find out, is trying to log in with credentials of a different provider. If you know somebody who has BT or the likes, you can actually do that. If they "log off" their PPPoE session and you fire in their details on your line, you can get a connection to another provider with that. (we've tested it with BT login credentials on an IBB dsl line, so I know that works).

The line speed won't change, as that's fixed in the DSLAM, but the trunk at the ISP side and where the Internet comes from.

/Martin

SOLACY

nuxxx wrote: »

Do you happen to no the IP address of the portloaise server by any chance?

It's 86.whatever.whatever.whatever

Lucutus

I know the topic of the thread is Eircom, however, my dealings over the past 8 weeks with BT Ireland technical support have ended in them laying the blame at Eircom's door... :rolleyes:

...Here's my ping graph for the last two days, all idle time:



It starts at around 16:30, ramps up to 400+ms at 18:00 and then starts to drop again at around 23:00 finally resting back to an acceptable level after midnight.

This seems too uniform to be congestion, or equipment that simply can't handle load.

If I was to guess, and I totally am, I'd suggest it's the result of some sort of profile or throttling of the connection during preset times of the day.

Can anyone 'in the know' shed some light onto the actual reasons behind this type of service disruption?

zugvogel

Lucutus wrote: »

I know the topic of the thread is Eircom, however, my dealings over the past 8 weeks with BT Ireland technical support have ended in them laying the blame at Eircom's door... :rolleyes:

...Here's my ping graph for the last two days, all idle time:



It starts at around 16:30, ramps up to 400+ms at 18:00 and then starts to drop again at around 23:00 finally resting back to an acceptable level after midnight.

This seems too uniform to be congestion, or equipment that simply can't handle load.

If I was to guess, and I totally am, I'd suggest it's the result of some sort of profile or throttling of the connection during preset times of the day.

Can anyone 'in the know' shed some light onto the actual reasons behind this type of service disruption?

Lucutus I'm in Kildare too, and see pretty much the same as yourself. Funny thing is though in the evening if I ping speedtest.airwire.ie (west side of the country Galway) the result is 40ms pings while boards.ie and other servers which I presume are on the East coast gives 450ms.

What software are you using to produce the graphs?

Zug

Marlow

zugvogel wrote: »

Lucutus I'm in Kildare too, and see pretty much the same as yourself. Funny thing is though in the evening if I ping speedtest.airwire.ie (west side of the country Galway) the result is 40ms pings while boards.ie and other servers which I presume are on the East coast gives 450ms.

What software are you using to produce the graphs?

Zug

Hi Zug,

the software he is using pingplotter. You can download that yourself.

As for your assumptions, they are actually wrong, but what you see is interesting.

The way this countries infrastructure is put together everything kind of ties together in Dublin. Well, not everything.

Airwire for example:
- if you are connected to Smart, your traffic will travel to us directly to Galway within Smarts network, the same applies to UPC, because we don't have direct peering with them, they have however to Smart.
- if you aren't connected to Smart, then your traffic will come to us either via the INEX (given that your ISP exchanges there) in Dublin or via any of the other connections we have (Cogent, GBLX, Level3 via Dublin or Smart via Galway).

boards.ie is hosted in Digiwebs hosting facility, same scenario there, either it goes through the INEX or via their transit connections.

A traceroute to speedtest.airwire.ie will show you this. So if you try to do a traceroute to that and to www.boards.ie, when you see the pings ramping up, you might be able to identify, where it goes wrong for you.

/Martin

Lucutus

lo Zug,

I'm using the software from pingplotter.com, very useful for dealing with front line support to get over the 'network congestion' fob off, especially with a couple of days worth of data shown.

I'll try to ping that galway hostname later during the usual disruption and see what result I get.

If you don't mind, forward your details to me via pm and I'll put together a list of people in Kildare with these issues. We can all run the same tests and gather some meaningful information to use when dealing with ISPs and any other interested bodies.

Luc

Grommet

zugvogel wrote: »

Tracing route to boards.ie [89.234.66.107]

over a maximum of 30 hops:



1 12 ms 99 ms 99 ms ....my router ...

2 35 ms 39 ms 43 ms b-ras1.pgs.portlaoise.eircom.net [159.134.155.8]

3 419 ms 411 ms 427 ms 86.43.242.62

4 422 ms 430 ms 448 ms 86.43.244.153

5 433 ms 435 ms 443 ms to-inex-dub-deg-gw.digiweb.ie [193.242.111.20]

6 454 ms 455 ms 422 ms ip-89-234-66-107.dedi.digiweb.ie [89.234.66.107]

Trace complete.

Grommet my local exchange is Nurney and as you can see from the trace route the problem only starts from b-ras1.pgs.portlaoise.eircom.net [159.134.155.8] onwards not my local exchange - Nurney.

Heres the results of Irish ISP Speed Test and this is on a line that syncs at 5.2MBits/sec and is perfect during office hours!

Download speed 783 Kbps (socket test)
Upload speed 305 Kbps (socket test)
Quality of service 44 %
Maximum delay 470 ms
Round trip time 402 ms
Upstream jitter 0.4 ms
Upstream packet loss 0 %
Upstream packet order 100 %
Upstream discards 0 %

The description of the node that is adding the congestion indicates that it is the Broadband Aggregation Server (BRAS) which is used to terminate the PPOE connection to your modem. I asked some questions of my mates in eircom about this and was told that they have noticed a large increase in traffic on one of the links on that BRAS over the last few months and that it started to congest a few weeks ago. So they have a new link being put in next week that should immediately alleviate the situation. I'll see if I can get an update next week and if so I'll post it here.

G.

Grommet

PogMoThoin wrote: »

:rolleyes: You sure bout that? How do You know this? Do You work for Eircom?

I sit and wait, I'm getting this in the mean time as upto 7.6mbit:

Pog - I'm told the upgrade for the link that connects to Frenchpark was upgraded a couple of nights ago. Have you seen any improvement in your service since then?

Hank Scorpio

i got a call today from eircom saying that there aware of the issues with some of there servers routing me and it will be fixed, but theres no time frame or estimate when it will be well worth calling me..

smellslikeshoes

nuxxx wrote: »

i got a call today from eircom saying that there aware of the issues with some of there servers routing me and it will be fixed, but theres no time frame or estimate when it will be well worth calling me..

One benefit about an answer like that is it shows that there is someone with something between their ears there. Theres enough vacuous idiots in Eircom support giving dates that they pulled out of their ass for things.

Hank Scorpio

i dont believe it but my connection appears to be fixed.

First time in 3 months its been stable at this time.

Praise the lord!:pac:

edit: I see alot of other people complaining ATM in other threads? Eircom must be doing some server work? Can anyone confirm this?

johnplayerblue

i can't confirm it nuxx but i will say that there better be some ball sack sucking coming my way from eircom after the night i've had.

neilk32

nuxxx wrote: »

i dont believe it but my connection appears to be fixed.

First time in 3 months its been stable at this time.

Praise the lord!:pac:

edit: I see alot of other people complaining ATM in other threads? Eircom must be doing some server work? Can anyone confirm this?

Ya its been good for me on xbl but it appears to be slow loading web pages ah well enjoy it while we can

PogMoThoin

Grommet wrote: »

Pog - I'm told the upgrade for the link that connects to Frenchpark was upgraded a couple of nights ago. Have you seen any improvement in your service since then?

Wow, it appears to be fixed, no sign of any contention this evening



Thanks for the info

Hank Scorpio

lets hope it lasts pog..

Iv got a feeling its because eircoms dns servers were down tonight for alot of people is the reason why we didnt experience our usual problems, 2moro will tell the tale, but ye its looking good!

Prisoner6409

I have been having similar issues and while the stated download speeds from speedtest and the like appeared to have recovered tonight web pages are taking forever to load. I did use the "tracert" command and got the following result:

1 <1ms <1ms <1ms 192.168.1.254
2 * * * Request Timed Out
3 <8ms <8ms <8ms ge8-0-0.coreb.cwm.dublin.eircom.net [86.43.243.2]

and so on. What I would like to know is what significiance the first hop(2) has. When I used pingplotter at 2(the first hop I suspect) it showed "100% packet Loss" at this stage. Can someone explain what this actually means and what might be going on here. It does not matter what site I ping it's always the same "100% packet loss" at stage 2.

vorbis^cs

OMG got a phone call from eircom compliants yesterday and today all day internet working great for gaming bit slow loadin web pages or not at all but i only wnated it for the games happy days lets hope im not speaking to soon

java

Lucutus wrote: »

I know the topic of the thread is Eircom, however, my dealings over the past 8 weeks with BT Ireland technical support have ended in them laying the blame at Eircom's door... :rolleyes:

How can BT lay the blame with eircom when your traces clearly route through the BT network? Am I missing something?

Lucutus

java wrote: »

How can BT lay the blame with eircom when your traces clearly route through the BT network? Am I missing something?

I know, the whole reason I used ftp.esat.net was to keep it in their network and demonstrate that the issue wasn't in INEX or some other transit carrier.

I received a helpful and informative email in mid-december which was forwarded from front line support in BT:

Unnamed BT Network Engineer wrote:

From looking at the traceroutes the high latency occurs from the very
first hop. What this shows us is that the latency is occurring somewhere
between the customers DSL modem and the BT bas. The only equipment on
the BT side between our bas and the customer is a circuit into the Eircom
network. Our Eircom BECS circuits have ample bandwidth available to them
at peek times and have zero errors on them at any time. The bas boxes
themselves also have ample resource available and zero errors.

The email went on to absolve all BT network equipment and place the problem squarely between my DSL router and the first hop, BT's BECS circuits.

I've tested tracerts with a couple of other local broadband customers (Eircom users) during the problematic peak times, and they have the same issue, even though it's not apparent to them with their usage patterns, (webmail and boards.ie browsing etc.).

Something else that's interesting, is that I can turn off my router, wait a short time, restart it, and I have roughly a 1 in 10 chance to get a different route through the network, with a 12ms stable ping. This connection gives me only 4 hops to ftp.esat.net instead of 8.

The saga continues...

Grommet

Lucutus wrote: »

The email went on to absolve all BT network equipment and place the problem squarely between my DSL router and the first hop, BT's BECS circuits.
..

This is more or less what I have been told by my techhie friends in eircom....

The eircom BECS service goes through the same BRAS node that eircom Retail services use - only difference is the customer connections are tunnelled through the BRAS and not terminated there as with the eircom service. So, since there is congestion on one of the paths out of the BRAS in Portlaoise the BT bitstream traffic will also hit this. The reason sometimes you get a good connection is because there is a second path out of the BRAS that apparently is not as congested and your traffic may route this way if you get assigned an address from a different subnet.

As I mentioned before, the congestion is to be fixed with new transmission links that are due at the end of next week. Should fix the issue for you.

NullZer0

OP - how much do you know about networking/internetworks ?