 rebus9
join:2002-03-26
Tampa Bay |
rebus9
Member
2024-Jun-23 8:16 pm
DHCP Lease Is Only 10 Minutes??Can anyone think of a valid reason why Frontier (broadband fiber) would only issue DHCP leases for 10 minutes? ——————————————————————————————————————————————- show dhcp client binding interface ge-0/0/0 detail
Client Interface/Id: ge-0/0/0.0
Hardware Address: 10:39:e9:**:**:**
State: BOUND(LOCAL_CLIENT_STATE_BOUND)
Lease Expires: 2024-06-23 19:26:31 EDT
Lease Expires in: 566 seconds
Lease Start: 2024-06-23 19:16:31 EDT
Server Identifier: 47.205.**.*
Client IP Address: 47.205.**.**
Update Server No
DHCP options:
Name: dhcp-lease-time, Value: 10 minutes
——————————————————————————————————————————————- Charter issues for 24 hours at a time. show dhcp client binding interface ge-0/0/7 detail
Client Interface/Id: ge-0/0/7.0
Hardware Address: 10:39:e9:**:**:**
State: BOUND(LOCAL_CLIENT_STATE_BOUND)
Lease Expires: 2024-06-24 05:16:56 EDT
Lease Expires in: 35852 seconds
Lease Start: 2024-06-23 09:42:32 EDT
Server Identifier: 72.31.**.*
Client IP Address: 72.185.**.**
Update Server No
DHCP options:
Name: dhcp-lease-time, Value: 19 hours, 34 minutes, 24 seconds
|
|
| |
Yours are 10 minutes? Mine are 30. |
|
| |
Drhoo
Member
2024-Jun-23 11:46 pm
how do I check this in linux mint? |
|
|
ARRIS NVG468MQ
Greenwave FiOS-G1100
Cisco 7941/7961
|
said by Drhoo:how do I check this in linux mint? this is something you would need to check in your router if you are going through a router |
|
bennor
Premium Member
join:2006-07-22
New Haven, CT |
to rebus9
Probably your market/area. In my case my router which is direct connected to the ONT is indicating a WAN IP lease time of 30 minutes on 500/500 fiber. |
|
rebus9
join:2002-03-26
Tampa Bay |
to terabitz
said by terabitz:Yours are 10 minutes? Mine are 30. Sadly, yes. What drew my attention to it was that every day, around the same time of day, I loose all external connectivity for 3-5 minutes. When it happened yesterday, I logged in to the router to see if it corresponded to when the DHCP lease was renewed-- thinking maybe it was a combination of lease expiration and a sluggish DHCP server that required several requests over a period of minutes to issue the renewal. That's when I saw 10 minutes, as shown in the CLI output. The only reason I can think of for such a short lease, is if Frontier was doing maintenance on whatever issues DHCP, and needed another system to issue short-term leases from a different pool until the primary system was back online.... but I've had the same IP for over a month, so that tends to reduce the likelihood of that. |
|
ARRIS NVG468MQ
Greenwave FiOS-G1100
Cisco 7941/7961
|
Frontier is wonky with non-Frontier routers and DHCP leases for some reason. The only reason my Asus router tends to stay connected is because I cloned the MAC address of my NVG468MQ into it.
If I didn't have a cloned MAC address, my connection would drop every 2-4 days, and the only fix is a router reboot. |
|
 camperjust visiting this planet
Premium Member
join:2010-03-21
Bethel, CT |
to rebus9
said by rebus9:Can anyone think of a valid reason why Frontier (broadband fiber) would only issue DHCP leases for 10 minutes? When I had a Frontier DSL connection, the lease my router (OpenBSD running the ISC-DHCP client) saw was 10 minutes, but the DHCP server that gave out that lease was in the DSL modem. The modem apparently got a lease from Frontier and then had an intermediatary DHCP server in the modem give out the lease to my router. With that in mind, one reason I can think of for the DHCP server in the modem giving out a 10 minute lease would be so that the lease given to my router would react quickly if the modem got a new set of IP addresses from the Frontier main DHCP server(s). |
|
rebus9
join:2002-03-26
Tampa Bay |
to terabitz
said by terabitz:Frontier is wonky with non-Frontier routers and DHCP leases for some reason. The only reason my Asus router tends to stay connected is because I cloned the MAC address of my NVG468MQ into it.
If I didn't have a cloned MAC address, my connection would drop every 2-4 days, and the only fix is a router reboot. Eeks, I'm glad I don't have to go to that extreme. The connection dropping every day is only a recent phenomenon. I've been using my own router (always Juniper, and I'm on my 3rd one) continuously since I first subscribed when Verizon was the provider, back in 2008. If it was a 24 hour lease, dropping once a day might make some level of sense. But a 10 minute lease, compared to a once-a-day disconnect around the same time of the evening every day.... I'm still trying to connect those dots. |
|
|
| |
to terabitz
Is this DHCP issue with DSL? I've never had any issues with DHCP on my fiber connection (originally Verizon, now Frontier) and I've never used their router. |
|
ARRIS NVG468MQ
Greenwave FiOS-G1100
Cisco 7941/7961
|
said by Alphasite:Is this DHCP issue with DSL? I've never had any issues with DHCP on my fiber connection (originally Verizon, now Frontier) and I've never used their router. No, fiber. Might be an issue with Asus routers in particular. |
|
| |
to rebus9
My lease is always 30 minutes. Frontier fiber via Palo Alto firewall.
I've kept the same WAN IP address since I signed up for service.
:confizzled: |
|
 guppy_fish
Premium Member
join:2003-12-09
Palm Harbor, FL |
to terabitz
I have used for 6 years now a ASUS (RT-AC86U) AC2900, never has a single issue with Froniter Fiber DHCP.
Frontier has been lately aggressive in the lease renewal, that combined with Godaddy changing the terms of updating the DNS table using their API has been a pain. |
|
 aefstoggaflmOpen Source Fan
Premium Member
join:2002-03-04
Bethlehem, PA
ARRIS SB6141
|
to Drhoo
Know your Unix/Linux  There are more DHCP client programmes available. One, fairly frequent option is to use dhclient. It is usually set to write full "PACK" answer decoded into a file. File is usually somewhere in /var. Something like find /var -name '*leases*' can give you a hint. For ex /var/lib/dhcp/dhclient.leases Or give some detail about SW you are using. |
|
rebus9
join:2002-03-26
Tampa Bay |
to tired_runner
said by tired_runner:My lease is always 30 minutes. Frontier fiber via Palo Alto firewall.
I've kept the same WAN IP address since I signed up for service. Mine seems to change only around what I'm assuming is backend maintenance. I'll have the same IP for months, then it will change several times(*) within a day or two, and then settle in for several months again to a single IP. (*) Known, because every time the IP changes, I have to update my source IP in the remote endpoints of the 2 static VPN tunnels I have running. Annoying, but quick to fix. |
|
| |
I have same struggles as you |
|
camperjust visiting this planet
Premium Member
join:2010-03-21
Bethel, CT |
to Alphasite
said by Alphasite:Is this DHCP issue with DSL? Apologies for introducing DSL into the discussion. I did so because the fiber ~symptoms~ seemed to reflect exactly what I saw on DSL from Frontier. My guess is that Frontier has a similar DHCP infrastructure for fiber as they inherited from AT&T for DSL. Indeed, until Frontier provides a real dual-stack IPv4/IPv6 service, I'd proffer that the old AT&T DHCP infrastructure is basically still in place. |
|
| |
to aefstoggaflm
Mint I looked in there already and it's not in there. |
|
rebus9
join:2002-03-26
Tampa Bay |
to tired_runner
Are you getting daily connection dropouts, too? I could almost set my watch by when the dropout happens each day. Used to be around 10 a.m. daily, but has now moved to around 6 p.m. each evening. Lasts 3-5 minutes. |
|
 ·Frontier FiberOp..
|
If it's happening, I have yet to notice.
My tunnels have 24-hour lifetimes. Maybe I'll keep a closer eye.
I leverage automated re-routing to backup cable modem connection via state monitoring on the Palo Alto firewall connected to Frontier. When the Palo can't ping Google, it takes about 60 seconds to drop the default route. |
|
rebus9
join:2002-03-26
Tampa Bay |
rebus9
Member
2024-Jun-25 5:13 pm
said by tired_runner:If it's happening, I have yet to notice.
My tunnels have 24-hour lifetimes. Maybe I'll keep a closer eye.
I leverage automated re-routing to backup cable modem connection via state monitoring on the Palo Alto firewall connected to Frontier. When the Palo can't ping Google, it takes about 60 seconds to drop the default route. For me, the duration is just enough to trigger WhatsUp notifications (set for a 3 minute timer before alert). And if we're watching TV, usually the stream won't have enough in the buffer to last the duration so we spend a minute or so watching the spinning donut of death until the connection comes back up. There's a JUNOS script the router can run to detect outages and alter the default (0.0.0.0/0) route, but I'd rather have the awareness that it's happening. One of these days, I'll find enough motivation to call Frontier tech support and wade through the quicksand that is Level I tech support. <shudder> They'll tell me to reboot my PC, reboot my router, and all the other stuff written on their canned script. Finally, they will blame my "enterprise branch office" (Juniper's term, not mine) router and say I should be using their kiddie router. Been there, done that, don't have the patience for it anymore. I work with the Frontier enterprise team at $DAYJOB (we have Enterprise DIA Fiber at various locations) and they're pretty sharp folks, but the residential Level I team is just.... painful. |
|
·Frontier FiberOp..
|
said by rebus9:One of these days, I'll find enough motivation to call Frontier tech support and wade through the quicksand that is Level I tech support. Now you understand why I simply failover to cable modem.  One time my Frontier connection dropped completely, and stood this way. The only reason I noticed is because YouTube TV asked to verify my location, and I became curious to see what my current public address was. I opened a ticket on their app, and their system was able to tell my connection was down. Tech visit was scheduled quickly. To cut a long story short, Frontier tech noticed significant light level drop to my ONT and ended up re-terminating the strand to my handoff inside my apartment. It's been rock solid since. I'm not suggesting you try hacking your fiber strand in order to get their attention, but...... |
|
rebus9
join:2002-03-26
Tampa Bay |
rebus9
Member
2024-Jun-25 5:41 pm
said by tired_runner:I'm not suggesting you try hacking your fiber strand in order to get their attention, but...... The thought of partially unseating the fiber at the ONT, just enough to cause degredation so a work ticket is generated (then re-seat it after the call), had occurred to me but I didn't want a tech chasing a phantom issue when the root cause could be quite different. To rule it out, when this first started I pulled the fiber connector and properly cleaned the face and ferrule just to rule it out. No joy. And just to rule out a physical interface problem, when I get the motivation I'm going to move the Frontier connection to another interface. At the end of a long workday, usually the last thing I feel like doing is reprovisioning an interface-- especially one terminiating VPN tunnels. I love and use the JUNOS 'replace' command, but it's not the whole solution when swapping interfaces. |
|
| |
You can check to see if you're in a maintenance window: » wholesale.frontier.com/t ··· ntenance |
|
rebus9
join:2002-03-26
Tampa Bay |
rebus9
Member
2024-Jun-27 8:30 pm
Thanks, I'll keep that link handy.
Well it happened. Today the connection went down hard, and has not come back. Rebooted ONT, replaced the ethernet cable between ONT and router, even tested the same router interface with another ISP connection and it worked 100%, so it's known good.
Unfortunately, today is only Thursday and the soonest truck roll available isn't until next week. Thank goodness for the backup Spectrum internet connection. (cheap insurance when your whole household telecommutes)
|
|
| rebus9 |
rebus9
Member
2024-Jul-1 1:20 pm
Tech was hear a few hours early. Cool guy, like most of the Frontier guys in the area are, and has been doing this for over 30 years.
Anyway, he said when he looked at my ticket and the model of the ONT, he knew right away it had to be the ONT. Said the Alcatel GPON units are kind of junky, so he swapped the ONT and ran down the street to the splitter to move me to different fiber for "XPON". I'm assuming that's XGS-PON because he said the ONT he installed is what they're using right now for the 5 Gbps tier.
Who am I to argue, right? So now if I want to upgrade to 2 or 5 gigabit, it's just a phone call with no truck roll needed.
|
|
| |
If it's the black one it's the FRX523. |
|
rebus9
join:2002-03-26
Tampa Bay
1 edit |
rebus9
Member
2024-Jul-1 2:37 pm
Yup, that would be it. It's larger than the one taken out, so the tech had to do some field engineering to the inside of the enclosure with his lineman pliers to make it fit.
Oddly, speeds have been consistently lower by 5-10 Mbps up/down since the swap, compared to what they were before. I wonder if it's just more subscribers on the XGS-PON line card, versus the GPON one I came from. I used to get consistent mid to upper 940s, and now I'm getting 930s. I'll run another test late tonight out of curiosity.
UPDATE: I've run speedtests periodically over the past 5 hours, to servers I know always give me consistent mid to upper 940s performance (942-948 is typical). Since the ONT swap and move to XGS-PON, not a single server is giving me more than low 930s. Obviously 930 is still fast, but I sure am curious why the sudden loss of 10-15 Mbps when theoretically I'm on fiber that is designed for higher performance than GPON. Hmmm.
|
|
ARRIS NVG468MQ
Greenwave FiOS-G1100
Cisco 7941/7961
|
You're lucky that you got 940 up/down with a GPON ONT. When I first got gig, I still had my O-211M-H ONT and it got ~500 down and 940 up. Had a tech out for slow speeds after awhile, and he gave me a FOG421. Since then I've only gotten anywhere from 840 to 910 down and 940 up. |
|
| |
to rebus9
Now you have your excuse to get this: » www.speedtest.net/my-res ··· 963bd570I get 2372.16 / 2375.46 on my i9-14900KS |
|
