Damaged tar message after upgrade

[RickNY]

I upgraded my system from 19.7 to 20.1, I believe - on Friday last week.  Last night around 3:30 AM, it looks like the firewall rebooted for whatever reason and got stuck with these repeating "Damaged tar" messages.  I tried rebooting a couple more times with the same results.. I discovered that pressing CTRL-C allowed it to continue the boot and the firewall came back up.  I did one or two more reboots - but still got stuck with the damaged tar messages that required CTRL-C to finish booting.  I was getting ready to clean install and restore from a config backup when I rebooted one more time from the GUI, and this time it made it through the entire boot without issue. 

Any way I can figure out what this was/is?  I'm running on a Netgate RCC-VE2440 x64/serial with an Intel 30GB SSD.


Thanks
Rick

[franco]

Hi Rick,

Do you have pictures or captures of the context of the message?

It's hard to go on a message that is not emitted by our code without context around where it was executed.


Cheers,
Franco

[RickNY]

So, what I ended up doing was taking a full config backup, and then doing a clean install of 20.1 on my RCC-VE 2440.. One of the things that stood out was that even when I got the firewall back up and running, it was taking a very long time to initiate a reboot, which typically completes in about 60-90 seconds.  After this happened, it was taking 3-4 minutes for a reboot.  Much of that time was spent during the shutdown process -- if I issued a reboot command from the GUI, the webpage would reload before the router even started rebooted.

I cant be 100% certain, but if I recall, the messages I was seeing about the damaged tar were showing up right when the router was attempting to start flowd_aggregate or flowd on the reboot.
tar: Damaged tar archive
    tar: Retrying...
    tar: Damaged tar archive
    tar: Retrying...
    tar: Damaged tar archive
    tar: Retrying...

Looking back at syslog messages from when this started around 3:20 AM on 2/4, it looks like my ISP remotely rebooted my modem, there was a WAN IP change via DHCP, and dpinger started complaining constantly about the gateway being down, and when I woke up around 6:00 AM, there was no internet connection and the serial console was just spitting out those tar messages.  It looks like dpinger didn't pick up the new gateway to monitor. 

Is it possible that archived logs from flowd became corrupt or something?  Why didn't dpinger change the gateway monitoring address after the WAN IP change?  Is it possible this could be related to the similar issue with flowd reported here?  https://forum.opnsense.org/index.php?topic=15698.0 -- if so, should I apply that patch?

In the end, I think the clean install and config restore was best.  My ISP does not change WAN IPs very often, maybe 2 or 3 times a year.  Now a reboot is taking about 60 seconds.

[franco]

Oh, ok, go to System: Settings: Miscellaneous and set "disable" for all periodic backups.


Cheers,
Franco