diff options
| author | Mike Christie <michaelc@cs.wisc.edu> | 2007-08-27 02:41:25 -0500 |
|---|---|---|
| committer | Mike Christie <michaelc@cs.wisc.edu> | 2007-08-27 19:37:54 -0500 |
| commit | 13b02660d0866725d10aa3bba03b949df23fef20 (patch) | |
| tree | ce6ecc82d309838701e9aaa5ebb721bb49010f4c /etc | |
| parent | a3e4c5202918409d54a893216a1d6555dc38cfe9 (diff) | |
| download | open-iscsi-13b02660d0866725d10aa3bba03b949df23fef20.tar.gz | |
mv nop handling to kernel
During root boot and shutdown the target could send us nops.
At this time iscsid cannot be running, so the target will drop
the session and the boot or shutdown will hang. This patch
moves the nop handling to the kernel.
Diffstat (limited to 'etc')
| -rw-r--r-- | etc/iscsid.conf | 19 |
1 files changed, 7 insertions, 12 deletions
diff --git a/etc/iscsid.conf b/etc/iscsid.conf index e950e3f..ea05c89 100644 --- a/etc/iscsid.conf +++ b/etc/iscsid.conf @@ -83,18 +83,14 @@ node.conn[0].timeo.login_timeout = 15 # The value is in seconds and the default is 15 seconds. node.conn[0].timeo.logout_timeout = 15 -# To specify the intervale to send iSCSI Nop-outs as pings -# to the target edit this the line. -# The value is in seconds and the default is 10 seconds. -node.conn[0].timeo.noop_out_interval = 10 +# Time interval to wait for on connection before sending a ping. +node.conn[0].timeo.noop_out_interval = 5 # To specify the time to wait for a Nop-out response before failing # the connection, edit this line. Failing the connection will # cause IO to be failed back to the SCSI layer. If using dm-multipath # this will cause the IO to be failed to the multipath layer. -# The value is in seconds and the default is 15 seconds. -node.conn[0].timeo.noop_out_timeout = 15 - +node.conn[0].timeo.noop_out_timeout = 5 #****** # Retry @@ -212,11 +208,10 @@ discovery.sendtargets.iscsi.MaxRecvDataSegmentLength = 32768 # Some targets like IET prefer after an initiator has sent a task # management function like an ABORT TASK or LOGICAL UNIT RESET, that # it does not respond to PDUs like R2Ts. To enable this behavior uncomment -# the following line (The default behavior is No): -# node.session.iscsi.FastAbort = Yes +# the following line (The default behavior is Yes): +node.session.iscsi.FastAbort = Yes # Some targets like Equalogic prefer that after an initiator has sent # a task management function like an ABORT TASK or LOGICAL UNIT RESET, that -# it continue to respond to R2Ts. To enable this uncomment this line (This -# is the default behavior): -node.session.iscsi.FastAbort = No +# it continue to respond to R2Ts. To enable this uncomment this line +# node.session.iscsi.FastAbort = No |