task abort: FAILED scmd

[degenhardt]

Good day. We have a nasty Proxmox kernel error on a RAID1 system: (out of /var/log/messages)

May 10 14:58:37 proxmoxXXX kernel: [57269.175789] scsi target0:1:0: volume handle(0x00be), volume wwid(0x05ac9c5b28c5dbfb)
May 10 14:59:06 proxmoxXXX kernel: [57298.871751] sd 0:1:0:0: attempting task abort! scmd(00000000c23993b1)
May 10 14:59:06 proxmoxXXX kernel: [57298.871758] sd 0:1:0:0: [sda] tag#4 CDB: Write(10) 2a 00 27 5d 21 68 00 00 10 00
May 10 14:59:06 proxmoxXXX kernel: [57298.871762] scsi target0:1:0: volume handle(0x00be), volume wwid(0x05ac9c5b28c5dbfb)
May 10 14:59:06 proxmoxXXX kernel: [57298.871767] sd 0:1:0:0: task abort: FAILED scmd(00000000c23993b1)

Our RAID-controller is: SAS2004 PCI-Express Fusion-MPT SAS-2 [Spitfire] (pretty old LSI low budget controller)
Our Proxmox system is: Virtual Environment 5.4-4 running on Debian 9

The problem seem to occur when the load average climbs over 5-7 (and higher).
Someone mentioned to me maybe upgrading the firmware of the RAID-card but I have not found any drivers yet.
Any ideas where to catch them or any other ideas on this issue ?
Thanks a bunch
Uwe