Is there any way to complain to someone in an actual position to affect change at CenturyLink about the subpar performance they deliver on IPv6? I am in the Portland Metro area, on GPON, in a 6rd only market (6rd gateway: 205.171.2.64) and performance is just atrocious. Transfers to/from Google Drive regularly timeout or get stuck at T-1 speeds, my IPv6 enabled media services (YouTube TV and Netflix) regularly buffer, sometimes v6 webpages fail to load, etc.
The samples below are better than what I usually see, if I were to take these ping tests in a few hours when we hit peak Internet usage they'll be coming back in the 200 to 300ms range on IPv6 while v4 remains perfectly stable.
Not sure what to do here, I can't use a Hurricane Electric prefix (thanks Netflix) and I don't really want to abandon IPv6, but what choice do I have? I expect some overhead with 6rd vs. dual stack but these performance experiences suggest to me that CenturyLink is in need of an upgrade on our 6rd gateway at the very least, and possibly better IPv6 peering on top of that. :(
% ping -c 5 205.171.2.64PING 205.171.2.64 (205.171.2.64): 56 data bytes64 bytes from 205.171.2.64: icmp_seq=0 ttl=250 time=298.492 ms64 bytes from 205.171.2.64: icmp_seq=1 ttl=250 time=45.663 ms64 bytes from 205.171.2.64: icmp_seq=2 ttl=250 time=37.306 ms64 bytes from 205.171.2.64: icmp_seq=3 ttl=250 time=98.971 ms64 bytes from 205.171.2.64: icmp_seq=4 ttl=250 time=59.707 ms --- 205.171.2.64 ping statistics ---5 packets transmitted, 5 packets received, 0.0% packet lossround-trip min/avg/max/stddev = 37.306/108.028/298.492/97.555 ms% traceroute 205.171.2.64traceroute to 205.171.2.64 (205.171.2.64), 64 hops max, 52 byte packets 1 172.31.255.254 (172.31.255.254) 2.695 ms 3.255 ms 2.151 ms 2 ptld-dsl-gw49.ptld.qwest.net (207.225.84.49) 5.915 ms 5.851 ms 5.515 ms 3 ptld-agw1.inet.qwest.net (207.225.86.129) 3.888 ms 4.075 ms 4.936 ms 4 tuk-edge-13.inet.qwest.net (67.14.4.206) 7.559 ms 8.631 ms 12.120 ms 5 * tuk-rdbr-01.inet.qwest.net (205.171.11.22) 283.584 ms *%
% ping -c 5 8.8.8.8PING 8.8.8.8 (8.8.8.8): 56 data bytes64 bytes from 8.8.8.8: icmp_seq=0 ttl=117 time=7.195 ms64 bytes from 8.8.8.8: icmp_seq=1 ttl=117 time=7.309 ms64 bytes from 8.8.8.8: icmp_seq=2 ttl=117 time=7.053 ms64 bytes from 8.8.8.8: icmp_seq=3 ttl=117 time=14.846 ms64 bytes from 8.8.8.8: icmp_seq=4 ttl=117 time=7.505 ms --- 8.8.8.8 ping statistics ---5 packets transmitted, 5 packets received, 0.0% packet lossround-trip min/avg/max/stddev = 7.053/8.782/14.846/3.036 ms% ping6 -c 5 2001:4860:4860::8888PING6(56=40+8+8 bytes) [deleted my ip] --> 2001:4860:4860::888816 bytes from 2001:4860:4860::8888, icmp_seq=0 hlim=117 time=42.824 ms16 bytes from 2001:4860:4860::8888, icmp_seq=1 hlim=117 time=53.860 ms16 bytes from 2001:4860:4860::8888, icmp_seq=2 hlim=117 time=30.458 ms16 bytes from 2001:4860:4860::8888, icmp_seq=3 hlim=117 time=51.041 ms16 bytes from 2001:4860:4860::8888, icmp_seq=4 hlim=117 time=35.029 ms --- 2001:4860:4860::8888 ping6 statistics ---5 packets transmitted, 5 packets received, 0.0% packet lossround-trip min/avg/max/std-dev = 30.458/42.642/53.860/8.976 ms%
↧
IPv6 Performance Issues
↧
[Qwest] two new DSLAMS in neighborhood, two streets adjacent missed
Hi! I am looking for someone who may work for CLINK/LUMEN that can help. A fiber expansion was complete in my neighborhood. They added two new DSLAMs which are now closer to peoples homes. Many near the new DSLAMs can now order faster speeds due to the units being installed and turned-up. However, two streets adjacent to the DSLAMs still report 3Mbps speeds if you try to order service. It appears CLINK/LUMEN did measure the Cu distance on these roads, as they placed orange markers along these roads, just like they placed along all of the other roads. I am having a hell of a time finding someone who can see if the database is still being updated...or if these roads were just missed. Any ideas on who to contact....or are one of you an "insider" that can help? thx.
↧
↧
C3000A firmware upgrade
I am showing a firmware upgrade is available CAD004-31.165L.13.BIN. However, when I attempt to download the firmware to my pc, I am taken to a centurylink page with an error message: Our apologies, this page is experiencing technical issues. Its been a few days. Has anyone been able to download this upgrade?
↧
[CenturyTel] Anyone here having major slowdonws and timeouts with CL internet
Something started yesterday morning with timeouts and very slow page-loads.
CL tech support, while friendly, had one heck of a long line waiting for TS and could only confirm the problem was not on my end and as far as they where concerned didn't exist.
Oh well, I moved on to better things to wok on then the net.
As of this afternoon it appears to be getting better, siill slow, but at least connecting eventually.
↧
[Qwest] WAN/PPPoE IP "static" or "dynamic" with Static IP Block/CBRAS?
I am getting CL "business" fiber at our "new" home (new for us, but actually 1930s), but am using the connection for home use. I did this so I could get static IPs and since I heard horror stories about residential static IPs and "Easy Pay".
I plan to get a static IP block (a /29) along with my GIgabit Fiber.
Is the WAN IP (the one assigned via PPPoE) "static" or "dynamic".
Also, if static, is the static IP assigned from my subnet, or is it a separate IP the subnet is routed to?
This is in a ex-Qwest market (Seattle, WA).
↧
↧
CenturyLink IPv6 6rd w/PPPoE and Third Party Equipment
Hey all,
Recently joined the GPON universe with Century Link in Portland. I switched to my own equipment with this setup:
ONT `- Managed Switch port tagged VID 201 `- Ubiquiti EdgeRouter-X w/PPPoE in same VLAN (untagged)
This setup seems to work well, I am getting 500-600Mbps down during peak hours and 800+ off peak (upstream is consistently around ~900Mbps so they do seem to have some downstream contention issues here but not enough to make an issue out of).
I resigned myself to not having provider IPv6 but recently I’ve stumbled across some old interweb posts referencing a 6rd setup. None mention actual settings, they’re all configured on CL equipment, and none say which markets this is supported in.
Anyone here have any insight? The tech who did my install let slip they are moving this market to straight Ethernet/DCHP and assumes we’ll have native IPv6 when that happens but he wasn’t privy to any timeline and it could presumably be months or years.
My next plan would be a Hurricane Electric tunnel but if I could get 6rd working with CL that’s better, IMHO, would prefer not to deal with Netflix or other CDNs who wrongly think HE tunnels are proxies.
↧
C3000A and C3000Z routers
I signed up for Century Link internet November 2020. C3000A was purchased; landline in basement now has a splitter (?) for DSL plug-in and the landline plug-in. Wifi drops occasionally. Many calls to CL for help. They now provided C3000Z to replace C3000A. Same problems: when connecting, it looks like wifi is connected, but then typing a website, the wifi-icon shows 'warning' and then shows "no internet". Many times, I unplug laptops (happens on both laptops) and I walk to top of basement stairs, then walk down a few-or-all-steps to see if I can get a signal and reconnect. Yesterday, I walked all the way downstairs, waved my laptop around and around trying to connect, a few feet from the C3000Z. (other times, I have to power down laptop and start all over again.) CL tech this week-end said to buy an extender. The C3000Z is only about 10-12 feet from kitchen table/laptop, with the floor/ceiling obviously in between basement and kitchen. Any ideas why I am having sooooo many connection problems?
02-18-2021 My two notes back to those who replied don't seem to appear. Anyway, I just took the laptop downstairs (router is now hanging from the ceiling right under the kitchen table), tried connecting to wifi 5 feet from the router, and same thing, it connected and dropped within 10 seconds "no internet, secure", and then about 30 seconds later automagically re-connected. Haven't gone to B.B. to look at extenders. Thanks again to the helpful suggestions; one thing at a time.
↧
[Qwest] Fiber: PPPoE not connecting on OPNsense or C3000A
I recently got Fiber service in Seattle, WA. I have put in my PPPoE credentials into my OPNsense box, and while it worked initially, but eventually stopped connecting.
If I put back the C3000A gateway I was given, it does not connect either.
The ONT shows green Network and Ethernet lights, and my Ethernet port seems to work.
My OPNsense logs give something like "DSL service disabled" and PPPoE disconnects afterwards. This cycle repeats.
The account still has a "pending" status. Is this to blame?
I have unplugged the ONT for now and will wait until the morning to plug it back in. Would this work?
I don't want to use Comcast if I can avoid it (I want the 1000 Mbps upload), but maybe I'll have no option and will just have to pray for uploads beyond 35 Mbps
↧
Henderson NV - need recommendation on modem
Hi Team,
Installing CL Internet service for a HOA/Camera solution using CenturyLink DSL service. I went to look at the install and found CL dropped the line (green line with a RG-11 connector, cable labled with DSL, but no modem. I asked the HOA manager about the modem, and she did not order a modem.
question: Can I just source a "approved" modem/router on my own? I really do not need/want a router, as I want to use my own router, but assume any combo modem/router can be put in bridge mode (read a few place that "transparent" terminology is used. Recommendations on model? Do not need wifi or will disable wifi if it has wifi.
Thanks in advance for any help!!
↧
↧
Change from Spectrum cable Internet to Centurylink DSL?
Around here Spectrum's lowest tier (available to me) is the 200Mbps package. That's way more bandwidth (and cost, and tired of the "try to get a new promo" thing) than I need at a 2nd location so thinking maybe to drop Spectrum and go with Centurylink DSL with their $49 for life deal.
I currently have Centurylink POTS and when I put in my address for availability it says up to 60Mbps.
Does that give any indication as to the type of DSL connection?
Any "gotchas" maybe I'm unaware of in consideration of such a change from cable modem to DSL?
↧
[Qwest] first i've heard of this, anyone else?
Features of the Zyxel C3510XZ
WiFi 6 technology with dual-band 2.4 GHz and 5 GHz 802.11 ax (backwards compatible with 802.11 a/b/g/n/ac)
Auto channel selection
Up to 4 wireless networks (SSIDs)
WPS / WPS 2.0 compatible
IPv4 / IPv6 compatible
WPA2-PSK security
https://www.centurylink.com/home/help/internet/modems-and-routers/zyxel-c3510xz.html
↧
Greenwave C4000BG review
Decided to upgrade my bonded VDSL2 connection with a new 4000BG that was recently released.
First some background. I am 1/2 mile from the DSLAM with 60/5 service for the last few years. Previously using a C3000A for about 2 years. Lines here are old but have been solid with no disconnects for a long time. Noticed a few months ago I started picking up more FEC errors and a couple of CRC errors every couple of days but no disconnects or performance hits. SNR on both lines were between 9-10. FEC errors were about 2500-3000/hr on one line and 1500/2000/hr on the other line.
So in comes the C4000BG. OH MY what a big difference in the line numbers. Currently lines has been up for 46 days with with 0 CRC and averaging about 300-400 FEC errors/hr. Also the SNR improved to 16-17 on the lines. What a major different!
Can't say enough good things about the C4000BG. The only con I have run across is the upload on the C3000A was "smoother" when running speed tests on Speedtest.net or Centurylink's own test.
↧
Bugs and stability problems on Greenwave C4000XG
Howdy! I'm a new Centurylink customer, and am deeply disappointed with the C4000XG, and am surprised there aren't more discussions about the problems on here. My experience, and conversations on other forums, lead me to believe this is a deeply flawed device.
As of firmware CGX008-01.03.78.00, at least some (all?) multicast traffic is being dropped on all LAN interfaces, including WLAN. This breaks mDNS (bonjour), and automatic discovery for Airprint, Airplay, and all sorts of other things. (I've confirmed this personally with packet captures.)
I'm also seeing really surprising behavior which I've never seen from even the crappiest gateway - connections just die, like they're getting pushed out of state tables too early. Pretty common, but I'm seeing this on the LAN too, with non-idle connections!
eg, While typing in an SSH session to a device on the LAN, the connection just dies and I get disconnected. Feels like connection tracking or DPI run amok.
Getting Centurylink to investigate or acknowledge any of these issues is like pulling teeth. I would be shocked if I could even speak to someone there who knows what multicast is.
They've left this firmware in the wild for several weeks. I can't tell if they're just flailing, tired of supporting their own Taiwanese ODM device, or what. It's especially frustrating because as a fiber customer, they refuse to swap for any model of any other device.
Other opinions or experiences?
↧
↧
No Centurylink stores anymore?
I have a friend who is having billing issues with Centurylink internet. Getting nowhere with the 800 number. There used to be Centurylink stores where you could go in and go over billing issues in person. Are they all gone now? We are in Littleton, Colorado and all I can find when I search for a store is Walmart?
↧
Mini-Jumbo frame support on FTTH connections (?)
I recently have been playing with my PPPoE fiber connection (boredom) and stumbled across RFC4638. This allows 'oversized' packet sizes to your PPP node, to overcome the 1492 MTU limit (because of the 8 byte ppp overhead). I thought 'no way clink supports this'. Using pfsense, I was able to configure my PPPoE tunnel to force an MTU of 1500, thus putting my transfers to a maximum of 1508 bytes outbound. I also had to adjust the MTU of the bound interface to 1508. Lo and behold, it *actually* works (or at least, appears to work)... Here's my output from pinging with a 'normal' packet size of 1500:
ping -d -s 1472 www.cnn.com -4
PING turner-tls.map.fastly.net (151.101.53.67) 1472(1500) bytes of data.
1480 bytes from 151.101.53.67 (151.101.53.67): icmp_seq=1 ttl=58 time=2.68 ms
1480 bytes from 151.101.53.67 (151.101.53.67): icmp_seq=2 ttl=58 time=2.57 ms
1480 bytes from 151.101.53.67 (151.101.53.67): icmp_seq=3 ttl=58 time=2.96 ms
Note the end size of the packet is in the parenthesis (1500)... with default values this ping got a fragment detection warning.
Of course, I can't tell any improvements form this, but it is neat, nonetheless... has anyone else played with this / knows of a better way to confirm this?
This site has a nice overview of what I'm talking about:
https://kieran.ie/mtu-baby-jumbo-frames-and-fttc/
↧
Provisioning reduced on rural DSL, any options?
Last night my internet crapped out for a few hours due to a local outage, by the time I discovered it was a local outage a phone rep had already done online troubleshooting which found "service problems" it attempted to correct. Lo and behold this morning I woke up and the internet was back but my previously overprovisioned line has been reduced to match my speed, which results in my no longer getting the advertised speed and a significant hit to my downstream bandwidth. The connection had always been stable outside of the occasional local outage, so I'm not thrilled about the loss in speed on an already slow connection. Do I have any options in trying to get Centurylink to increase the provisioning again, or am I out of luck? I'm in a former Qwest area and read that line provisioning can improve automatically over time if the stats are good but I don't know if that applies when CTL intentionally reduced it.
↧
Bypassing CenturyLink modem with VLAN 201 tags still possible?
Hi, I'm in the Vancouver, WA area and have happily used my modem in bridge mode with my router doing PPPoE authentication.
Yesterday the WAN port on the modem died. The light on the front flashes between red/green and CenturyLink can't ping it. The ONT, I'm told, is working fine and there isn't a neighborhood outage. Neighbor has the same service and it's working so I tend to believe them.
So... they are sending a new modem but won't ship it without a tech appointment which will take **12 days**. 😕
Using Verizon Fios (in the NYC area) I was able to easily bypass the modem because they used ipoe.
I found a thread here in the forums and some blog posts that made it sound pretty simple to use VLAN tagging of the wan port with 201.
So I purchased a NetGear managed switch and followed all the directions but it's not working. The only thing I don't see on the switch is the ability to set the packet priority to 0.
Most of the posts about this are very old. Is it still possible?
Are they maybe using a different tag rather than 201?
Any help is greatly appreciated.
Screenshots attached.
I set port 7 as the tagged port (connected to the ONT) and port 8 untagged (connected to my router's WAN port).
[att=1]
[att=2]
[att=3]
Thanks!
↧
↧
why call in? /r
ok, it's Monday, during business hours (don't want to get off=shored!) call in to 800-244-1111, get the phone bot, play the game. wait 10 min, get one ring, and "hello, this is (unintelligible due to accent) how an i help you? i say "hello" and the call drops. call back on 877-299-0946, play the phone bot game again. wait 10 min, get a single ring (this time different MOH). a few minutes later, "hello, this is (unintelligible due to accent), how may i assist you?" i ask "are you in the US?" reply: "no. so you'd like to speak to someone in the US?" of course i said yes. still that different MOH, finally BACK TO THE PHONEBOT. play the game, wait 10 min, get a single ring and four minutes of dead air. p.s. the chat buttons on line on three different browsers don't work, and i've enabled all scripts to work on those pages. :mad: all i wanted to do was de-link my Directv acct from CTL (Directv calls it "disunion") so my DTV gets sent separately. (used to have a $5 discount for bundling thru Qwest) and DTV can't see my past billing because it's done thru CTL. /r off.
↧
IPv6 Performance Issues
Is there any way to complain to someone in an actual position to affect change at CenturyLink about the subpar performance they deliver on IPv6? I am in the Portland Metro area, on GPON, in a 6rd only market (6rd gateway: 205.171.2.64) and performance is just atrocious. Transfers to/from Google Drive regularly timeout or get stuck at T-1 speeds, my IPv6 enabled media services (YouTube TV and Netflix) regularly buffer, sometimes v6 webpages fail to load, etc.
The samples below are better than what I usually see, if I were to take these ping tests in a few hours when we hit peak Internet usage they'll be coming back in the 200 to 300ms range on IPv6 while v4 remains perfectly stable.
Not sure what to do here, I can't use a Hurricane Electric prefix (thanks Netflix) and I don't really want to abandon IPv6, but what choice do I have? I expect some overhead with 6rd vs. dual stack but these performance experiences suggest to me that CenturyLink is in need of an upgrade on our 6rd gateway at the very least, and possibly better IPv6 peering on top of that. :(
% ping -c 5 205.171.2.64PING 205.171.2.64 (205.171.2.64): 56 data bytes64 bytes from 205.171.2.64: icmp_seq=0 ttl=250 time=298.492 ms64 bytes from 205.171.2.64: icmp_seq=1 ttl=250 time=45.663 ms64 bytes from 205.171.2.64: icmp_seq=2 ttl=250 time=37.306 ms64 bytes from 205.171.2.64: icmp_seq=3 ttl=250 time=98.971 ms64 bytes from 205.171.2.64: icmp_seq=4 ttl=250 time=59.707 ms --- 205.171.2.64 ping statistics ---5 packets transmitted, 5 packets received, 0.0% packet lossround-trip min/avg/max/stddev = 37.306/108.028/298.492/97.555 ms% traceroute 205.171.2.64traceroute to 205.171.2.64 (205.171.2.64), 64 hops max, 52 byte packets 1 172.31.255.254 (172.31.255.254) 2.695 ms 3.255 ms 2.151 ms 2 ptld-dsl-gw49.ptld.qwest.net (207.225.84.49) 5.915 ms 5.851 ms 5.515 ms 3 ptld-agw1.inet.qwest.net (207.225.86.129) 3.888 ms 4.075 ms 4.936 ms 4 tuk-edge-13.inet.qwest.net (67.14.4.206) 7.559 ms 8.631 ms 12.120 ms 5 * tuk-rdbr-01.inet.qwest.net (205.171.11.22) 283.584 ms *%
% ping -c 5 8.8.8.8PING 8.8.8.8 (8.8.8.8): 56 data bytes64 bytes from 8.8.8.8: icmp_seq=0 ttl=117 time=7.195 ms64 bytes from 8.8.8.8: icmp_seq=1 ttl=117 time=7.309 ms64 bytes from 8.8.8.8: icmp_seq=2 ttl=117 time=7.053 ms64 bytes from 8.8.8.8: icmp_seq=3 ttl=117 time=14.846 ms64 bytes from 8.8.8.8: icmp_seq=4 ttl=117 time=7.505 ms --- 8.8.8.8 ping statistics ---5 packets transmitted, 5 packets received, 0.0% packet lossround-trip min/avg/max/stddev = 7.053/8.782/14.846/3.036 ms% ping6 -c 5 2001:4860:4860::8888PING6(56=40+8+8 bytes) [deleted my ip] --> 2001:4860:4860::888816 bytes from 2001:4860:4860::8888, icmp_seq=0 hlim=117 time=42.824 ms16 bytes from 2001:4860:4860::8888, icmp_seq=1 hlim=117 time=53.860 ms16 bytes from 2001:4860:4860::8888, icmp_seq=2 hlim=117 time=30.458 ms16 bytes from 2001:4860:4860::8888, icmp_seq=3 hlim=117 time=51.041 ms16 bytes from 2001:4860:4860::8888, icmp_seq=4 hlim=117 time=35.029 ms --- 2001:4860:4860::8888 ping6 statistics ---5 packets transmitted, 5 packets received, 0.0% packet lossround-trip min/avg/max/std-dev = 30.458/42.642/53.860/8.976 ms%
↧
CenturyLink IPv6 6rd w/PPPoE and Third Party Equipment
Hey all,
Recently joined the GPON universe with Century Link in Portland. I switched to my own equipment with this setup:
ONT `- Managed Switch port tagged VID 201 `- Ubiquiti EdgeRouter-X w/PPPoE in same VLAN (untagged)
This setup seems to work well, I am getting 500-600Mbps down during peak hours and 800+ off peak (upstream is consistently around ~900Mbps so they do seem to have some downstream contention issues here but not enough to make an issue out of).
I resigned myself to not having provider IPv6 but recently I’ve stumbled across some old interweb posts referencing a 6rd setup. None mention actual settings, they’re all configured on CL equipment, and none say which markets this is supported in.
Anyone here have any insight? The tech who did my install let slip they are moving this market to straight Ethernet/DCHP and assumes we’ll have native IPv6 when that happens but he wasn’t privy to any timeline and it could presumably be months or years.
My next plan would be a Hurricane Electric tunnel but if I could get 6rd working with CL that’s better, IMHO, would prefer not to deal with Netflix or other CDNs who wrongly think HE tunnels are proxies.
↧