Re: tmux, systemd and jack capture

Hey Thomas!

Thanks a lot for your input! You are totally right, I could have checked 
this myself. Works like a charm with screen. So I suppose I'll go with 
that solution - fantastic - also your script is very helpful! Kudos!

Cheers
Jan


Am 2019-04-25 15:41, schrieb Thomas Brand:
> On 2019-04-25 12:39, [email protected] wrote:
>> Hey Thomas,
>> 
>> systemctl --version
>> systemd 232
>> +PAM +AUDIT +SELINUX +IMA +APPARMOR +SMACK +SYSVINIT +UTMP
>> +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ +LZ4 +SECCOMP +BLKID +ELFUTILS
>> +KMOD +IDN
>> 
>> From that issue thread I see it's only in 240 - so this is not the
>> problem I suppose.
>> 
>> Cheers
>> Jan
>> 
>> 
>> Am 2019-04-25 12:33, schrieb Thomas Brand:
>>> On 2019-04-25 11:35, [email protected] wrote:
>>>> Hey!
>>>> 
>>>> I have a hard time setting up a tmux session started with systemd on
>>>> raspbian (patchbox OS, realtime kernel) to behaves like as if I
>>>> started my script manually.
>>> 
>>>> Cannot use real-time scheduling (RR/70)(1: Operation not permitted)
>>>> JackClient::AcquireSelfRealTime error
>>> 
>>>> 
> 
> It's yet unclear to me if this is related to jack at all.. Can you
> test this without jack? Basically leaving everything out except a
> script started by systemd that tries to lock memory or calls ulimit.
> 
> Similar to your tmux approach, with screen, here's a building block:
> 
> #!/bin/bash
> 
> start_in_detached_screen()
> {
>         #$1: username
>         #$2: screen name
>         #$3: command string (without ^M)
> 
>         su ${1} -c "screen -d -m -S \"$2\""
>         su ${1} -c "screen -S \"$2\" -p 0 -X stuff \"$3\"^M"
> }
> 
> start_in_detached_screen john_doe "my test" "cd /tmp && ls -ltra"
> 
> -> as root: start screen as given user and feed command to screen 
> session.
> It should work to start jackd as a specific user early in the boot
> process and see what happened by attaching the screen later on.
> 
> Greetings
> Thomas