[TriLUG] scp interrupted system call with copy from Ubuntu 18.04 sys to 20.04 sys

[Pete Soper via TriLUG]

After sending this it dawned on me that what I was really doing was 
copying a directory from machine "rick" to the Samba share residing on 
"morty" with an scp -r on "dl".

So I guess this is my first data point  in regards to Alan or somebody 
asking me to report how the new network storage is working. :-) I'm in 
an intermediate state, with morty (4GB Raspberry Pi 4 with a 2TB 
rotating drive with USB 3 interface) just sharing its big disk to the 
Linux and Windows clients in the house and being a caching DNS server. 
But every night morty copies all of the network storage to an identical 
2TB drive on rick while I get around to making the two drives a RAID 1. 
Up until today this had all been going without a hitch. The spin up time 
of the drive on morty is a bit faster than the drives on the old system 
(an Odroid C4) and performance is noticeably better, but this might be 
because I'd really abused an SSD on the previous system. And I finally 
went back and found some ssh tips you guys shared and got my keys sorted 
out so I can navigate between the various machines without aggravating 
the arthritis in my thumbs typing passwords. :-)

Cheers,
Pete

PS System "rick" is going to my daughter's place for her to use with her 
GIS grad studies after it gets a bit more prep with whatever name she 
wants. Now that I'm not dealing with the TI Eclipse-based IDE any more I 
only need a laptop, not something that will snap your neck doing a 
double take at how fast it is. This seems somehow apropos since she's 
the one who turned me on to the cartoons.


On 6/11/20 2:39 PM, Pete Soper wrote:
> Today on this system  "dl" running 20.04 (with all updates/upgrades) I 
> was doing an "scp -r rick:tms ." and got this and the copy is hung:
>
> /tms/vendor/BNO080/STM32 Nucleo-64/nucleo-f411re.pdf: truncate: 
> Interrupted system call
> nucleo-f411re.pdf                             100%   24MB 3.4MB/s   00:06
> ./tms/vendor/BNO080/STM32 Nucleo-64/nucleo-f411re.pdf: Interrupted 
> system call
> BNO080-Development-Kit-for-Nucleo-Quick-Start   0%    0 0.0KB/s --:-- ETA
>
> It's still in the hung state in case there's something I could do to 
> get more information. There is nothing in /var/log/syslog or kern.log. 
> Both machines are 64 bit Intel-based.
>
> What forum or place would I poke around to see if this is a known 
> problem and whether there's a patch for it and how do I frame my 
> search/question (e.g. what is the formal name of the chunk of Linux 
> this is likely to involve?).
>
> The network hardware hasn't been changed in years except a few weeks 
> ago I put a gigabit switch between three of the machines on wired 
> ethernet from the router and rick is one of them (longest cable run 
> about eight feet). Maybe I should run scp all night and see if I can 
> flush out more failures and then take the gigabit switch out and 
> repeat the experiment, or would this be a silly waste of effort?
>
> Thanks,
> Pete