I'm trying to get the crashdumps (to start getting them
automatically later) on an iSCSI-booted buster/sid, but after
"echo c > /proc/sysrq-trigger" all
I'm getting is the system hangup - no dump, but the system seems
to be locked - and no panic screen or whatever. Everything is frozen
up. However, I'm getting crashdumps when booting from local
disks. The problem is - my production hosts use iSCSI for
booting, so I really need to get these crashdumps whil on iSCSI.
I know that Linux isn't able to write the crashdump on "local" disk
when this "local" disk is iSCSI, so I've set up the NFS
crashdump resource - but to NFS activity is happening when I'm
triggering the panic.
Is there some trick that I'm unaware of ?