Can someone verify if:
- Connecting to an external PPTP servers works? Packets get stuck at the dd-wrt router and are not forwarded to the client machine
- Enabling logging on SPI works?
Cleared NVRAM and, even if some minor issues I've noticed were solved, this is still an issue.
Thanks
Just in case someone has a similar problem:
- you need to enable klogd (kernel logger) in order to see iptables logs (Services > System Log)
- you need to enable the conntrack helper to successfully do NAT for the GRE protocol (echo 1 > /proc/sys/net/netfilter/nf_conntrack_helper)
Joined: 06 Jun 2006 Posts: 7492 Location: Dresden, Germany
Posted: Sat May 27, 2017 9:15 Post subject:
flipy wrote:
flipy wrote:
Can someone verify if:
- Connecting to an external PPTP servers works? Packets get stuck at the dd-wrt router and are not forwarded to the client machine
- Enabling logging on SPI works?
Cleared NVRAM and, even if some minor issues I've noticed were solved, this is still an issue.
Thanks
Just in case someone has a similar problem:
- you need to enable klogd (kernel logger) in order to see iptables logs (Services > System Log)
- you need to enable the conntrack helper to successfully do NAT for the GRE protocol (echo 1 > /proc/sys/net/netfilter/nf_conntrack_helper)
i fixed this new behaviour of kernel 4.7 > in 32146 _________________ "So you tried to use the computer and it started smoking? Sounds like a Mac to me.." - Louis Rossmann https://www.youtube.com/watch?v=eL_5YDRWqGE&t=60s
Joined: 06 Jun 2006 Posts: 7492 Location: Dresden, Germany
Posted: Sat May 27, 2017 10:57 Post subject:
introduces driver update my marvell
jffs2 works again, but takes very long time to mount (nand driver still not bugfree)
covers the nf_conntrack_helper which is a kernel behaviour change since 4.7 _________________ "So you tried to use the computer and it started smoking? Sounds like a Mac to me.." - Louis Rossmann https://www.youtube.com/watch?v=eL_5YDRWqGE&t=60s
flashed successfuly 32149 on wrt3200acm.
for this i used "mtd -f /tmp/factory-to-ddwrt.bin mtd5" (so before update partition 2 was used).
after reboot i am on part 1 and i noticed that the setings remains and i checked mtd partitions and i see that nvram partition is still on 0x00000a900000-0x00000a940000, overlapping mtd7 (part 2).
is this the desired behavior ? i think that after an upgrade nvram should overlap the active linux partition.
Joined: 06 Jun 2006 Posts: 7492 Location: Dresden, Germany
Posted: Sat May 27, 2017 14:29 Post subject:
adi-brasov wrote:
flashed successfuly 32149 on wrt3200acm.
for this i used "mtd -f /tmp/factory-to-ddwrt.bin mtd5" (so before update partition 2 was used).
after reboot i am on part 1 and i noticed that the setings remains and i checked mtd partitions and i see that nvram partition is still on 0x00000a900000-0x00000a940000, overlapping mtd7 (part 2).
is this the desired behavior ? i think that after an upgrade nvram should overlap the active linux partition.
if you flash in a unusual way without changing the boot partition settings it will always boot from the same partition again. so you are responsible here what you're doing. the correct flashing method is using webflash or the "write" command. using the mtd command is custom and does not take care about any bootpartition settings. the nvram is not responsible for the bootpartition setting. just as a note. it just shows the state.
use ubootenv set boot_part 1 or 2 for selecting the bootpartition _________________ "So you tried to use the computer and it started smoking? Sounds like a Mac to me.." - Louis Rossmann https://www.youtube.com/watch?v=eL_5YDRWqGE&t=60s
Just upgraded to r32149 and speeds as ovpn client are very good once again. I’m going to and set it up to do some streaming and see how that goes. @ BS Thank You again!
The guest network DHCP lease time settings are in minutes. Default value of 3600 translates to 60 hrs.
Can it be changed to seconds in the next release?
flashed successfuly 32149 on wrt3200acm.
for this i used "mtd -f /tmp/factory-to-ddwrt.bin mtd5" (so before update partition 2 was used).
after reboot i am on part 1 and i noticed that the setings remains and i checked mtd partitions and i see that nvram partition is still on 0x00000a900000-0x00000a940000, overlapping mtd7 (part 2).
is this the desired behavior ? i think that after an upgrade nvram should overlap the active linux partition.
if you flash in a unusual way without changing the boot partition settings it will always boot from the same partition again. so you are responsible here what you're doing. the correct flashing method is using webflash or the "write" command. using the mtd command is custom and does not take care about any bootpartition settings. the nvram is not responsible for the bootpartition setting. just as a note. it just shows the state.
use ubootenv set boot_part 1 or 2 for selecting the bootpartition
Joined: 06 Jun 2006 Posts: 7492 Location: Dresden, Germany
Posted: Sat May 27, 2017 15:20 Post subject:
ad19 wrote:
The guest network DHCP lease time settings are in minutes. Default value of 3600 translates to 60 hrs.
Can it be changed to seconds in the next release?
sorry for last post. its in minutes doing in seconds makes no sense more me. i like leases with 14 days sometimes to track all users for a longer time. or lets say not track, but see how many where online. if i want to type in 14 days in seconds guess what how it looks: 1209600 seconds.
in addition consider that several buggy clients will disconnect on lease refresh. so using less than a minute does not make much sense or why do you want todo such crazy things? i mean a network with 1000 clients get a lot of traffic just with lease refreshes with such low values _________________ "So you tried to use the computer and it started smoking? Sounds like a Mac to me.." - Louis Rossmann https://www.youtube.com/watch?v=eL_5YDRWqGE&t=60s
The guest network DHCP lease time settings are in minutes. Default value of 3600 translates to 60 hrs.
Can it be changed to seconds in the next release?
sorry for last post. its in minutes doing in seconds makes no sense more me. i like leases with 14 days sometimes to track all users for a longer time. or lets say not track, but see how many where online. if i want to type in 14 days in seconds guess what how it looks: 1209600 seconds.
in addition consider that several buggy clients will disconnect on lease refresh. so using less than a minute does not make much sense or why do you want todo such crazy things? i mean a network with 1000 clients get a lot of traffic just with lease refreshes with such low values
No great reason to want it in seconds. Just for consistency, I guess. But it's a non-issue. Minutes work just fine as well.
Thank you.
Last edited by ad19 on Sat May 27, 2017 15:26; edited 1 time in total
Joined: 06 Jun 2006 Posts: 7492 Location: Dresden, Germany
Posted: Sat May 27, 2017 15:26 Post subject:
i changed it now to 1440 = 24 hours _________________ "So you tried to use the computer and it started smoking? Sounds like a Mac to me.." - Louis Rossmann https://www.youtube.com/watch?v=eL_5YDRWqGE&t=60s
