guest-image-tools: autoconfigure networking: hipersockets nic offline
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
nova-dpm |
Invalid
|
Undecided
|
Unassigned |
Bug Description
Started a DPM partition with 2 networks. One OSA and one Hipersockets backed network.
Both NICs got grouped correctly, but only the OSA NIC is "online".
Debug data:
# ip a
1: lo: mtu 65536 qdisc noqueue state UNKNOWN qlen 1
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enccw0.0.0005: mtu 1500 qdisc pfifo_fast st
ate UP qlen 1000
link/ether fa:16:3e:89:ed:f7 brd ff:ff:ff:ff:ff:ff
inet 192.168.223.7/24 brd 192.168.223.255 scope global dynamic enccw0.0.0005
valid_lft 86118sec preferred_lft 86118sec
inet6 fe80::f816:
valid_lft forever preferred_lft forever
# znetconf -u
Scanning for network devices...
# znetconf -c
Device IDs Type Card Type CHPID Drv. Name St
ate
-------
0.0.0002,
fline
0.0.0005,
line
# lschp
CHPID Vary Cfg. Type Cmg Shared PCHID
=======
0.01 1 2 11 - - 0164
0.05 1 1 11 2 1 0128
0.0f 1 2 25 - - 0118
0.17 1 2 11 - - 0174
0.1e 1 1 25 2 1 0131
0.2e 1 2 24 - - (07cb)
0.33 1 1 24 3 1 (07d0)
0.34 1 2 24 - - (07d1)
0.3a 1 2 24 - - (07d7)
# lsqeth
Cannot get device feature names: No such device
Device name :
-------
card_type : HiperSockets
cdev0 : 0.0.0002
cdev1 : 0.0.0003
cdev2 : 0.0.0004
chpid : 33
online : 0
portname : no portname required
portno : 0
route4 : n/a
route6 : n/a
state : DOWN
layer2 : 1
isolation : none
Device name : enccw0.0.0005
-------
card_type : OSD_10GIG
cdev0 : 0.0.0005
cdev1 : 0.0.0006
cdev2 : 0.0.0007
chpid : 05
online : 1
portno : 0
state : UP (LAN ONLINE)
layer2 : 1
isolation : none
# znetconf -r 0002 -n
Remove network device 0.0.0002 (0.0.0002,
Warning: this may affect network connectivity!
Successfully removed device 0.0.0002
# znetconf -a 0002
Scanning for network devices...
[ 1074.203074] qeth 0.0.0002: Completion Queueing supported
[ 1074.206762] qeth: register layer 3 discipline
[ 1074.207087] qeth 0.0.0002: Completion Queueing supported
[ 1084.200556] qeth 0.0.0002: Completion Queueing supported
[ 1094.200564] qeth 0.0.0002: Completion Queueing supported
[ 1104.200607] qeth 0.0.0002: Completion Queueing supporteD
[ 1114.200379] qeth 0.0.0002: The qeth device driver failed to recover an error
on the device
/sbin/znetconf: line 560: echo: write error: No such device
znetconf: Error: Failed to make 0.0.0002 online
# znetconf -c
Device IDs Type Card Type CHPID Drv. Name St
ate
-------
0.0.0002,
fline
0.0.0005,
line
# ls /sys/bus/
blkt hsuid portname state
buffer_count hw_trap portno subsystem
card_type if_name power switch_attrs
cdev0 inbuf_size priority_queueing uevent
cdev1 ipa_takeover recover ungroup
cdev2 isolation route4 vipa
chpid layer2 route6
driver online rxip
fake_broadcast performance_stats sniffer
# echo "1" > /sys/bus/
devices/
[ 1317.452472] qeth 0.0.0002: Completion Queueing supported
[ 1327.451296] qeth 0.0.0002: The qeth device driver failed to recover an error
on the device
[ 1327.451498] qeth 0.0.0002: Completion Queueing supported
[ 1337.450573] qeth 0.0.0002: Completion Queueing supported
[ 1347.450543] qeth 0.0.0002: Completion Queueing supported
[ 1357.450369] qeth 0.0.0002: The qeth device driver failed to recover an error
on the device
-bash: echo: write error: No such device
description: | updated |
Starting a new instance worked. I remember I've seen this in the past. The workaround was to delete and add the NIC via the HMC again... So this is not a problem of our dpm code - closing the issue.