Hi there
just ran into an issue with a failing device (ssd decided to die)
After rebooting the node, it won't bring up the interfaces, as the "ifupdown2-pre.service" wont succeed .. its basically a "/bin/udevadm settle" that waits till everything is okay.
But due to the failing disk device, udev won't settle.. and voila .. your network won't come up ...
manually deleting the device (echo 1 > /sys/block/sdc/device/delete) and restarting the services got me back a working system.
I think this behaviour should be fixed , as it prevents a recovery of a node in trouble by denying network connectivity :-/
just ran into an issue with a failing device (ssd decided to die)
After rebooting the node, it won't bring up the interfaces, as the "ifupdown2-pre.service" wont succeed .. its basically a "/bin/udevadm settle" that waits till everything is okay.
But due to the failing disk device, udev won't settle.. and voila .. your network won't come up ...
Code:
[ 1939.690975] blk_update_request: I/O error, dev sdc, sector 2000409088 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0
[ 1939.692053] Buffer I/O error on dev sdc, logical block 250051136, async page read
[ 1940.149921] sd 0:0:3:0: Power-on or device reset occurred
[ 1940.190951] mpt3sas_cm0: log_info(0x31120b10): originator(PL), code(0x12), sub_code(0x0b10)
[ 1940.190958] sd 0:0:3:0: [sdc] tag#2196 FAILED Result: hostbyte=DID_SOFT_ERROR driverbyte=DRIVER_OK
[ 1940.190960] sd 0:0:3:0: [sdc] tag#2196 CDB: Read(10) 28 00 77 3b d2 00 00 00 08 00
[ 1940.190962] blk_update_request: I/O error, dev sdc, sector 2000409088 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 0manually deleting the device (echo 1 > /sys/block/sdc/device/delete) and restarting the services got me back a working system.
I think this behaviour should be fixed , as it prevents a recovery of a node in trouble by denying network connectivity :-/