Boot issue with iSCSI target root device in SLES11

[Chandra Seetharaman]

Hello,

Installation completed fine.

Boot started fine, iSCSI daemon started, logged into iface, found the
storage.

While fsck'ing had a connection error, and immediately int started had
another connection error, and just hung there.

--------------
Waiting for device /dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2 to appear: ok
fsck 1.41.1 (01-Sep-2008)
[/sbin/fsck.ext3 (1) -- /] fsck.ext3 -a /dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2
/dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2: recovering journal
connection1:0: detected conn error (1011)
/dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2: clean, 154441/1310720 files, 1022000/5235168 blocks
fsck succeeded. Mounting root device read-write.
Mounting root /dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2
mount -o rw,acl,user_xattr -t ext3 /dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2 /root
kjournald starting. Commit interval 5 seconds
EXT3 FS on sdi2, internal journal
EXT3-fs: mounted filesystem with ordered data mode.
mount: can't find /root/proc in /etc/fstab or /etc/mtab
INIT: version 2.86 booting
iscsi: can not broadcast skb (-3)
connection1:0: detected conn error (1011)

--------------

Has anybody seen this problem ? Any suggestions ?

[Mike Christie]

Is this fsck is started from the init scripts during boot up? If so, the
reason for the hang is probably that iscsid is not up and cannot handle
the connection error. For fedora/RHEL, I think we did some hack where if
network storage is used we delayed the fsck from the init boot scripts.

Not sure about the reaosn for the conn error. Can you turn on debugging?

[Chandra Seetharaman]

Thanks for your help, Mike.

On Thu, 2009-10-29 at 13:06 -0500, Mike Christie wrote:
> Is this fsck is started from the init scripts during boot up? If so, the
> reason for the hang is probably that iscsid is not up and cannot handle
> the connection error. For fedora/RHEL, I think we did some hack where if
> network storage is used we delayed the fsck from the init boot scripts.

iscsi daemon was up... see the detail boot message below

>
> Not sure about the reaosn for the conn error. Can you turn on debugging?

how ? it is just the first reboot. Is there a kernel command line option
that I can specify ?

----------------------
Detailed boot message
-----------------------
Boot logging started on /dev/hvc0(/dev/console) at Mon Oct 26 21:24:47
2009

NET: Registered protocol family 17
[NETWORK] using interface eth0

[NETWORK] using static config based on
ip=9.3.103.95::9.3.103.1:255.255.255.0:9.3.103.95:eth0:none

Loading iSCSI transport class v2.0-870.
iscsi: registered transport (tcp)
FATAL: Module iscsi_ibft not found.

Starting iSCSI daemon

Logging in to [iface: default, target: iqn.1986-03.com.ibm:sn.118060811,
portal: 9.3.111.253,3260]

eth0: link up, 10Gbps, full-duplex
scsi1 : iSCSI Initiator over TCP/IP
scsi 1:0:0:0: Direct-Access NETAPP LUN 0.2 PQ: 0 ANSI: 4
sd 1:0:0:0: [sde] 41943040 512-byte hardware sectors: (21.4GB/20.0GiB)
sd 1:0:0:0: [sde] Write Protect is off
sd 1:0:0:0: [sde] Write cache: disabled, read cache: enabled, doesn't support
DPO or FUA
sd 1:0:0:0: [sde] 41943040 512-byte hardware sectors: (21.4GB/20.0GiB)
sd 1:0:0:0: [sde] Write Protect is off
sd 1:0:0:0: [sde] Write cache: disabled, read cache: enabled, doesn't support
DPO or FUA
sde: sde1 sde2
sd 1:0:0:0: [sde] Attached SCSI disk
scsi 1:0:0:1: Direct-Access NETAPP LUN 0.2 PQ: 0 ANSI: 4
sd 1:0:0:1: [sdf] 41943040 512-byte hardware sectors: (21.4GB/20.0GiB)
sd 1:0:0:1: [sdf] Write Protect is off
sd 1:0:0:1: [sdf] Write cache: disabled, read cache: enabled, doesn't support
DPO or FUA
sd 1:0:0:1: [sdf] 41943040 512-byte hardware sectors: (21.4GB/20.0GiB)
sd 1:0:0:1: [sdf] Write Protect is off
sd 1:0:0:1: [sdf] Write cache: disabled, read cache: enabled, doesn't support
DPO or FUA
sdf: sdf1 sdf2
sd 1:0:0:1: [sdf] Attached SCSI disk
scsi 1:0:0:2: Direct-Access NETAPP LUN 0.2 PQ: 0 ANSI: 4
sd 1:0:0:2: [sdg] 41943040 512-byte hardware sectors: (21.4GB/20.0GiB)
sd 1:0:0:2: [sdg] Write Protect is off
sd 1:0:0:2: [sdg] Write cache: disabled, read cache: enabled, doesn't support
DPO or FUA
sd 1:0:0:2: [sdg] 41943040 512-byte hardware sectors: (21.4GB/20.0GiB)
sd 1:0:0:2: [sdg] Write Protect is off
sd 1:0:0:2: [sdg] Write cache: disabled, read cache: enabled, doesn't support
DPO or FUA
sdg: sdg1 sdg2
sd 1:0:0:2: [sdg] Attached SCSI disk
scsi 1:0:0:3: Direct-Access NETAPP LUN 0.2 PQ: 0 ANSI: 4
sd 1:0:0:3: [sdh] 41943040 512-byte hardware sectors: (21.4GB/20.0GiB)
sd 1:0:0:3: [sdh] Write Protect is off
sd 1:0:0:3: [sdh] Write cache: disabled, read cache: enabled, doesn't support
DPO or FUA
sd 1:0:0:3: [sdh] 41943040 512-byte hardware sectors: (21.4GB/20.0GiB)
sd 1:0:0:3: [sdh] Write Protect is off
sd 1:0:0:3: [sdh] Write cache: disabled, read cache: enabled, doesn't support
DPO or FUA
sdh: sdh1 sdh2
sd 1:0:0:3: [sdh] Attached SCSI disk
scsi 1:0:0:4: Direct-Access NETAPP LUN 0.2 PQ: 0 ANSI: 4
sd 1:0:0:4: [sdi] 41943040 512-byte hardware sectors: (21.4GB/20.0GiB)
sd 1:0:0:4: [sdi] Write Protect is off
sd 1:0:0:4: [sdi] Write cache: disabled, read cache: enabled, doesn't support
DPO or FUA
sd 1:0:0:4: [sdi] 41943040 512-byte hardware sectors: (21.4GB/20.0GiB)
sd 1:0:0:4: [sdi] Write Protect is off
sd 1:0:0:4: [sdi] Write cache: disabled, read cache: enabled, doesn't support
DPO or FUA
sdi: sdi1 sdi2
sd 1:0:0:4: [sdi] Attached SCSI disk
Login to [iface: default, target: iqn.1986-03.com.ibm:sn.XXXXXX, portal:
X.X.XXX.XXX,3260]: successful

Waiting for device /dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2
to appear: ok

fsck 1.41.1 (01-Sep-2008)

[/sbin/fsck.ext3 (1) -- /] fsck.ext3 -a
/dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2

/dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2: recovering
journal

connection1:0: detected conn error (1011)
/dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2: clean,
154441/1310720 files, 1022000/5235168 blocks

fsck succeeded. Mounting root device read-write.

Mounting root /dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2

mount -o rw,acl,user_xattr -t ext3
/dev/disk/by-id/scsi-360a98000486e5a76464a53614a394461-part2 /root

kjournald starting. Commit interval 5 seconds
EXT3 FS on sdi2, internal journal
EXT3-fs: mounted filesystem with ordered data mode.
mount: can't find /root/proc in /etc/fstab or /etc/mtab

INIT: version 2.86 booting

iscsi: can not broadcast skb (-3)
connection1:0: detected conn error (1011)

-------------
>
> >