not sure if you have debug information enabled, but can you try to run
'pstack' on the two processes at the time they get hanged, and send the
output? "pstack -p <pid">"
note that you won't be able to run pstack on them if you run them under
strace.
--guy
Erik Burrows wrote:
I'm running into trouble running basic iscsiadm commands, which seems
to stem from a communication deadlock between iscsiadm and iscsid, but
more likely a problem with my configuration. Your assistance will be
greatly appreciated.
Initiator System:
Gentoo Linux
Kernel version: 2.6.32-gentoo-r7
open-iscsi version: 2.0.871.3
Target System:
Ubuntu Linux
Kernel version: 2.6.31-21-generic
iscsitarget version: 0.4.16+svn
Initiator /etc/iscsi/initiatorname.iscsi:
InitiatorName=iqn.2005-03.com.erikburrows:two
InitiatorAlias=two
Initiator /etc/iscsid.conf unchanged from default:
node.startup = automatic
node.session.timeo.replacement_timeout = 120
node.conn[0].timeo.login_timeout = 15
node.conn[0].timeo.logout_timeout = 15
node.conn[0].timeo.noop_out_interval = 5
node.conn[0].timeo.noop_out_timeout = 5
node.session.err_timeo.abort_timeout = 15
node.session.err_timeo.lu_reset_timeout = 20
node.session.initial_login_retry_max = 4
node.session.cmds_max = 128
node.session.queue_depth = 32
node.session.iscsi.InitialR2T = No
node.session.iscsi.ImmediateData = Yes
node.session.iscsi.FirstBurstLength = 262144
node.session.iscsi.MaxBurstLength = 16776192
node.conn[0].iscsi.MaxRecvDataSegmentLength = 131072
discovery.sendtargets.iscsi.MaxRecvDataSegmentLength = 32768
node.session.iscsi.FastAbort = Yes
Target /etc/ietd.conf:
Target iqn.2010-04.com.erikburrows:griffin.var.lun1
Lun 0 Path=/var/iscsi/lun1.img,Type=fileio
Alias LUN1
The problem is that when I try to do "discovery", or manual create/
login, the scsiadm command on the initiator host just hangs:
r...@two:/etc/iscsi# iscsiadm -m discovery -d 8 -t st -p 192.168.1.4 -
P 1
iscsiadm: ip 192.168.1.4, port 3260, tgpt -1
iscsiadm: Max file limits 1024 1024
^Ciscsiadm: caught SIGINT, exiting...
Same result when manually creating the node, then logging in.
An strace of the iscsiadm command shows a hang on a recv() from the
socket to the iscsid process:
open("/sys/class/iscsi_host", O_RDONLY|O_NONBLOCK|O_LARGEFILE|
O_DIRECTORY|0x80000) = 3
fcntl64(3, F_GETFD) = 0x1 (flags FD_CLOEXEC)
getdents(3, /* 2 entries */, 32768) = 32
getdents(3, /* 0 entries */, 32768) = 0
close(3) = 0
socket(PF_FILE, SOCK_STREAM, 0) = 3
connect(3, {sa_family=AF_FILE, pa...@iscsiadm_abstract_namespace},
110) = 0
write(3, "\n
\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
5244) = 5244
recv(3, ^C <unfinished ...>
An strace of the iscsid process shows a hang on a read() from the same
socket:
munmap(0xb7873000, 4096) = 0
read(8, "\n
\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
5248) = 5244
read(8, ^C <unfinished ...>
Process 3917 detached
This looks like a communication deadlock to me, but is it really a mis-
configuration on my part?
Thanks,
Erik Burrows
--
You received this message because you are subscribed to the Google Groups
"open-iscsi" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/open-iscsi?hl=en.
