Final 6.2.1 update causing LAN instability/sluggishness

Hmm, seems I'm not done yet either :(

Moved the RN516 temporarily, so I could check the messages on the display. After the OS re-install it worked. But after shutdown and moving it back near the router, on restart, the same issue starts occuring. Looks like I am going to have to revert to the RC3 also, if I want to permantely fix this.

I downgraded my RN312 from the "final Final" to 6.2.1 RC3.
This fixed the problem.

The RC3 release comes with ConnMan 1.25.
The final Final 6.2.1 came with ConnMan 1.27.

ConnMan 1.27 was released Saturday, December 13th, 2014.
With this version, the ConnMan daemon (connmand) on my RN312 took about 99.5 % CPU and it sent out a huge number of TX packets over eth0.

Wild guess: connmand 1.27 floods routers with tons of faulty traffic.
Perhaps it has not been tested adequate on the final Final 6.2.1.

Not sure if this is related, seems similar:
viewtopic.php?f=64&t=77397

Is it possible to downgrade the firmwware without accessing the GUI and without loosing data?

Using ssh or remotely.
Perhaps contact support, especially if you are within 90 days of purchase.

_frantic_ wrote:
The RC3 release comes with ConnMan 1.25. The final Final 6.2.1 came with ConnMan 1.27.

If reverting to RC3 has solved the issue for those of us with this issue, this ConnMan discrepancy would seem to be at the root of the problem. Maybe it doesn't play well with IPv6? Or a certain router chipset?

Hope the OS 6 devs are paying attention.

Yes, _frantic_'s "wild guess" seems very plausible :) And it would explain the symptoms I see.

I put a continuous ping on the router from my PC. With the RN connected, it cannot reach the router. As soon as you disconnect the RN, the router connection comes back.

But how do you downgrade to RC3?

I don't have a working connection to RN516 because as soon as it is connected to the router all connections disappear. So I cannot reach the admin interface, nor reach it over SSH. Any suggestions?

@dsnpevl: I have connected both ReadyNAS and PC to a switch. The router is connected to the same switch. In this constellation, PC and ReadyNAS can talk to each other even when the router goes down.


Sent from my iPad using Tapatalk HD

OK, seems you must definitely disable all portforwarding as well. Then I managed to re-install the OS and have the web admin and SSH connection back.
One other weird thing I noticed, is while I only connected on LAN1, the NIC for LAN2 also show up in the router. This seems to be related to an an observation made here:

mountaindude wrote:
After restarting connamn (or the entire NAS) I have THREE IPs responding to pings: 192.168.1.16, 192.168.17 and 192.168.106 What??

Anyway, I can now try some stuff to get it to work. Logged into SSH and followed this procedure to check for connman packages (command line version).

root@nas-XX-XX-XX:~# apt-cache showpkg connman
Package: connman
Versions:
1.27-netgear1 (/var/lib/apt/lists/apt.readynas.com_packages_readynasos_dists_6.2.1_main_binary-amd64_Packages) (/var/lib/dpkg/status)
 Description Language:
                 File: /var/lib/apt/lists/apt.readynas.com_packages_readynasos_dists_6.2.1_updates_binary-amd64_Packages
                  MD5: cb5a9cf686ae7cfd884a7a436fa64503

1.26-netgear1 (/var/lib/apt/lists/apt.readynas.com_packages_readynasos_dists_6.2.1_updates_binary-amd64_Packages)
 Description Language:
                 File: /var/lib/apt/lists/apt.readynas.com_packages_readynasos_dists_6.2.1_updates_binary-amd64_Packages
                  MD5: cb5a9cf686ae7cfd884a7a436fa64503

1.0-1.1+wheezy1+b1 (/var/lib/apt/lists/mirrors.kernel.org_debian_dists_wheezy_main_binary-amd64_Packages)
 Description Language:
                 File: /var/lib/apt/lists/apt.readynas.com_packages_readynasos_dists_6.2.1_updates_binary-amd64_Packages
                  MD5: cb5a9cf686ae7cfd884a7a436fa64503


Reverse Depends:
  connman-dbg,connman 1.27-netgear1
  connman-dbg,connman 1.26-netgear1
Dependencies:
1.27-netgear1 - iptables (0 (null)) libc6 (2 2.15) libdbus-1-3 (2 1.1.1) libglib2.0-0 (2 2.31.8) libreadline6 (2 6.0) libxtables7 (0 (null)) dbus (0 (null)) lsb-base (0 (null)) bluez (0 (null)) wpasupplicant (0 (null))
1.26-netgear1 - iptables (0 (null)) libc6 (2 2.15) libdbus-1-3 (2 1.1.1) libglib2.0-0 (2 2.31.8) libreadline6 (2 6.0) libxtables7 (0 (null)) dbus (0 (null)) lsb-base (0 (null)) bluez (0 (null)) wpasupplicant (0 (null))
1.0-1.1+wheezy1+b1 - iptables (0 (null)) libc6 (2 2.9) libdbus-1-3 (2 1.1.1) libglib2.0-0 (2 2.28.0) libgnutls26 (2 2.12.17-0) libxtables7 (0 (null)) dbus (0 (null)) lsb-base (0 (null)) bluez (0 (null)) wpasupplicant (0 (null))
Provides:
1.27-netgear1 -
1.26-netgear1 -
1.0-1.1+wheezy1+b1 -
Reverse Provides:

So this means I can try to downgrade to the 1.26-netgear1 version, hoping it doesn't have the flaw of the 1.27.

root@nas-XX-XX-XX:~# sudo apt-get install connman=1.26-netgear1
Reading package lists... Done
Building dependency tree
Reading state information... Done
Recommended packages:
  bluez wpasupplicant
The following packages will be DOWNGRADED:
  connman
0 upgraded, 0 newly installed, 1 downgraded, 0 to remove and 51 not upgraded.
Need to get 383 kB of archives.
After this operation, 4,096 B disk space will be freed.
Do you want to continue? [Y/n] Y
Get:1 http://apt.readynas.com/packages/readynasos/ 6.2.1/updates connman amd64 1.26-netgear1 [383 kB]
Fetched 383 kB in 1s (251 kB/s)
debconf: delaying package configuration, since apt-utils is not installed
dpkg: warning: downgrading connman from 1.27-netgear1 to 1.26-netgear1
(Reading database ... 32568 files and directories currently installed.)
Preparing to replace connman 1.27-netgear1 (using .../connman_1.26-netgear1_amd64.deb) ...
Unpacking replacement connman ...
Setting up connman (1.26-netgear1) ...
insserv: warning: current stop runlevel(s) (empty) of script `connman' overrides LSB defaults (0 1 6).
insserv: can not symlink(../init.d/nut-server, ../rc2.d/S03nut-server): File exists
insserv: can not symlink(../init.d/forked-daapd, ../rc2.d/S05forked-daapd): File exists
insserv: can not symlink(../init.d/avahi-daemon, ../rc2.d/S04avahi-daemon): File exists
insserv: can not symlink(../init.d/dbus, ../rc2.d/S03dbus): File exists
insserv: can not symlink(../init.d/apache2, ../rc2.d/S02apache2): File exists
insserv: can not symlink(../init.d/lsyncd, ../rc2.d/S01lsyncd): File exists
insserv: can not symlink(../init.d/mdadm, ../rc2.d/S03mdadm): File exists
insserv: can not symlink(../init.d/connman, ../rc2.d/S04connman): File exists
insserv: can not symlink(../init.d/rsync, ../rc2.d/S03rsync): File exists
insserv: can not symlink(../init.d/cron, ../rc2.d/S03cron): File exists
insserv: can not symlink(../init.d/nut-server, ../rc3.d/S03nut-server): File exists
insserv: can not symlink(../init.d/forked-daapd, ../rc3.d/S05forked-daapd): File exists
insserv: can not symlink(../init.d/avahi-daemon, ../rc3.d/S04avahi-daemon): File exists
insserv: can not symlink(../init.d/dbus, ../rc3.d/S03dbus): File exists
insserv: can not symlink(../init.d/apache2, ../rc3.d/S02apache2): File exists
insserv: can not symlink(../init.d/lsyncd, ../rc3.d/S01lsyncd): File exists
insserv: can not symlink(../init.d/mdadm, ../rc3.d/S03mdadm): File exists
insserv: can not symlink(../init.d/connman, ../rc3.d/S04connman): File exists
insserv: can not symlink(../init.d/rsync, ../rc3.d/S03rsync): File exists
insserv: can not symlink(../init.d/cron, ../rc3.d/S03cron): File exists
insserv: can not symlink(../init.d/nut-server, ../rc4.d/S03nut-server): File exists
insserv: can not symlink(../init.d/forked-daapd, ../rc4.d/S05forked-daapd): File exists
insserv: can not symlink(../init.d/avahi-daemon, ../rc4.d/S04avahi-daemon): File exists
insserv: can not symlink(../init.d/dbus, ../rc4.d/S03dbus): File exists
insserv: can not symlink(../init.d/apache2, ../rc4.d/S02apache2): File exists
insserv: can not symlink(../init.d/lsyncd, ../rc4.d/S01lsyncd): File exists
insserv: can not symlink(../init.d/mdadm, ../rc4.d/S03mdadm): File exists
insserv: can not symlink(../init.d/connman, ../rc4.d/S04connman): File exists
insserv: can not symlink(../init.d/rsync, ../rc4.d/S03rsync): File exists
insserv: can not symlink(../init.d/cron, ../rc4.d/S03cron): File exists
insserv: can not symlink(../init.d/nut-server, ../rc5.d/S03nut-server): File exists
insserv: can not symlink(../init.d/forked-daapd, ../rc5.d/S05forked-daapd): File exists
insserv: can not symlink(../init.d/avahi-daemon, ../rc5.d/S04avahi-daemon): File exists
insserv: can not symlink(../init.d/dbus, ../rc5.d/S03dbus): File exists
insserv: can not symlink(../init.d/apache2, ../rc5.d/S02apache2): File exists
insserv: can not symlink(../init.d/lsyncd, ../rc5.d/S01lsyncd): File exists
insserv: can not symlink(../init.d/mdadm, ../rc5.d/S03mdadm): File exists
insserv: can not symlink(../init.d/connman, ../rc5.d/S04connman): File exists
insserv: can not symlink(../init.d/rsync, ../rc5.d/S03rsync): File exists
insserv: can not symlink(../init.d/cron, ../rc5.d/S03cron): File exists
[ ok ] Restarting connman (via systemctl): connman.service.

So now the big question: will it still work after restart. Since I have connection still, I used the web admin to restart. Will let you know how this goes.

_frantic_ wrote:
@dsnpevl: I have connected both ReadyNAS and PC to a switch. The router is connected to the same switch. In this constellation, PC and ReadyNAS can talk to each other even when the router goes down.

Yes, tried that too, but for me when the RN516 is on, no matter where in the LAN, it disables the router anyway. Then the IP addresses no longer get resolved and I cannot access anything from my PC. I now have the PC on the switch and the switch on the router. The RN is directly on another port of the router.

Trying downgrading to 1.26 now.
Update1: first signs are promissing for the downgrade to connman 1.26 on final OS 6.1.2. Will try some shutdown/restart next to see if it keeps working. Also: the weird second listing of the LAN2 NIC in the router has disappeared now.
Update2: After shutdown and power-off and power-on, the connection still works. Will try restart from web admin next.
Update3: Takes a very long time to reboot. The web admin is telling me: "Connection lost. Connection could not be restored. Please reboot your device manually and try again". But actually, the RN was still rebooting (it says "booting..." in the front display). When you wait long enough, the connection comes back and no router issues. The web gui needs to learn to have a little patience...
Update4: Actually remebered the issues with the port forwarding, so I powered down the RN (again) and re-instated my port forwardings on the router. After restart the network stays alive and I can access from the web admin. Next I will shut down/power down and move the RN back to it's opertional posistion. If that works, the fix above has been proven sound, from my perspective.
Update5: I moved the RN back near the switch and restarted it. Connections are still working and my website fotografeer.nl is up-and-running again, finally. I consider the fix decribed above a success. Thanks to all for starting threads on this issue. And many thanks to _frantic_ for finding the root cause. Netgear should be forever in our debt for doing their work for them :)

So this fix works for me:
- Look for the installed versions of connman:

apt-cache showpkg connman

- Provided 1.26-netgear1 is listed, you can then downgrade from 1.27-netgear1 to 1.26-netgear1

sudo apt-get install connman=1.26-netgear1

- Details and full screen output above.

Enjoy your holidays and have a happy new year with many trouble-free OS updates!