New release and bugs
-
1/ No info to debug.
2/ Sounds like faulty HW.
3/ WFM
4/ Restarting the packages is by design and required. (No info to debug "many instances".)
5/ ??? (Pretty much useless, using undisciplined local clock for NTP -> you could just as well shut the service down.)
6/ WFM.
7/ Are you actually having an Intel-based WiFi card? Does not sound so per #8, so why are you messing with that at all.
8/ Bug FreeBSD guys. -
4/ Restarting the packages is by design and required. (No info to debug "many instances".)
Off topic, but it shouldn't be required. I've had the same DHCP IP for a couple of years now… If apinger thinks its down, even if its not... every package restarts... It should check if the IP has changed, if has; restart packages... ok rant is over =D
-
Are all those bugs which have been reported in Redmine? If they are then they should have been assigned a target release for the fix so you should be able to see if they will be fixed in 2.2 or 2.2.X etc.
Most of those don't seem like critical bugs so they're unlikely to hold up a 2.2 release. For example the fact that packages can end up being restarted multiple times as boot is not ideal but I know of only one package that is affected by that and even then only in a specific configuration.
Steve
-
altq on em0 priq queue { default, medium, hi } queue default on em0 priority 1 priq ( red , ecn , default ) queue medium on em0 priority 5 priq ( ecn ) queue hi on em0 priority 7 priq ( ecn ) altq on bridge0 priq queue { default, medium, hi } queue default on bridge0 priority 1 priq ( ecn , default ) queue medium on bridge0 priority 5 priq ( ecn ) queue hi on bridge0 priority 7 priq ( ecn )PS: LAN is bridge0.
-
-
When you get fsck on each reboot then your system does not shutdown correctly in the first place. Other than that, creating a monster thread for totally unrelated stuff won't go anywhere, just end up with huge mess. Move your separate issues into proper forums and provide some info.
-
4/ Restarting the packages is by design and required. (No info to debug "many instances".)
Off topic, but it shouldn't be required. I've had the same DHCP IP for a couple of years now… If apinger thinks its down, even if its not... every package restarts... It should check if the IP has changed, if has; restart packages... ok rant is over =D
I have DHCPv6 and static IPv4 on WAN. And SLAAC address to, because can't be disabled.
After some time i have tens copy of arpwatch in memory for example. -
…
Other than that, creating a monster thread for totally unrelated stuff won't go anywhere, just end up with huge mess. Move your separate issues into proper forums and provide some info.Agree, my mistake.
PS: But all of these problems disappear if I go back to 2.1.5.
-
4/ Restarting the packages is by design and required. (No info to debug "many instances".)
Off topic, but it shouldn't be required. I've had the same DHCP IP for a couple of years now… If apinger thinks its down, even if its not... every package restarts... It should check if the IP has changed, if has; restart packages... ok rant is over =D
Several services will break if they don't re-bind to the IP address even if it's the same as it was before.
-
For bridge you have to set the sysctl where pf filters will be applied by default they are on member interfaces.
-
1. The queues not working on interface bridge.
I see you started a separate thread for that, which is better. Keep the details there.
2. “fsck” runs 3 times on every reboot.
fsck only runs three times if mount returns errors every time.
Run #1 only happens if the drive could not be mounted without error.
Run #2 only happens if mount still did not complete without error after run #1
Run #3 only happens if mount still did not complete without error after run #2So the problem isn't necessarily that fsck isn't showing an error, it's that mount does show an error. You still may have filesystem problems.
I can only get fsck to run if I force an unclean shutdown, and even then, it only happens once.
Due to this, many of the other problems become suspicious. Source files could be getting misread or may be corrupted in some way. Please take a backup, and wipe+reload the install, perhaps using a fresh disk in the process.
3. In the GUI filter logs IPv6 is absent, but present in logfile.
I see a separate thread for this now, too. Though it seems to work fine here and anywhere else I've used it.
4. Some services can start many instances on the every obtain the new leases for WAN (on every restart packages).
Needs more specific detail in a separate thread. (Which services? Under what conditions? What log messages appear when it happens?)
5. Not working services->NTP->Orphan mode – settings are ignoring.
The NTP code got a huge update contributed by some forum member(s). There is a dedicated thread just for the new/updated NTP code and if you call attention to this, they can probably fix it pretty easily. Though it may not warrant holding up 2.2, 2.2.1 should follow up before too long after.
6. Static routes not bring up after firmware update (update packages), only after new reboot.
Does the package reinstall actually complete properly? I could see this happening If the package install got hung up. I cannot reproduce this. Static routes are still present after a firmware upgrade even with packages reinstalling, but they all succeed.
7. Can't agree with the license, set legal.intel_ipw.license_ack=1 in /boot/loader.conf.local do nothing.
This works fine for me, either in /boot/loader.conf.local or in /boot/loader.conf - make sure you're using the right tunable. The only one I see at boot time is
legal.intel_iwi.license_ack=1Which isn't what you listed.
8. WiFi settings. Setting “802.11ng” on my “Atheros 9285” the card is not working at all. But setting “802.11g” the card is working and working in “ng” mode.
There is another thread for 802.11n on atheros. There are people who have it working on certain cards, but it may require some other options to be set such as specific channels, WME, etc. It's here on the 2.2 board somewhere.
-
