Hi all, I will describe the problem I saw step by step. Please take patience.
I'm using tgt as iscsi target server (fujita/tgt from github) on Ubuntu 16.04. iscsi initiator runs the stock open-iscsi coming with Ubuntu 16.04. The default "/etc/iscsi/iscsid.conf" defines "node.session.queue_depth = 32". I changed it to 128, and run "sudo iscsid restart". Then I log in to an iscsi target. However, " iscsiadm -m node -p xxx -T yyy" shows my iscsi session still uses "queue_depth" as 32, not 128. Then I run "iscsiadm -m node -p xxx -T yyy --op update -n node.session.queue_depth -v 128" to forcefully change the value. This time "iscsiadm -m node" shows the correct "128" value I set. Now I run fio benchmark with a large queue size (1024), and keeps monitoring iscsi target. I had a patch in tgt to report the current live commands received on a connection. To my surprise, this max "queued_cmd" is always 32, not 128, no matter how hard I push fio queue size. I was expecting that, since I set initiator session.queue_depth to 128, the target should receive a max 128 cmds in its queue. I want the initiator to push as many requests as possible to target to get more bandwidth. However the target only receives up to 32 outstanding cmds from initiator. Why doesn't initiator push more? Is my understanding about "node.session.queue_depth" correct? -Shawn -- You received this message because you are subscribed to the Google Groups "open-iscsi" group. To unsubscribe from this group and stop receiving emails from it, send an email to open-iscsi+unsubscr...@googlegroups.com. To post to this group, send email to open-iscsi@googlegroups.com. Visit this group at https://groups.google.com/group/open-iscsi. For more options, visit https://groups.google.com/d/optout.
