open-iscsi Google Group

Re: Information about iSCSI pings that almost timed out

erezzi.l...@gmail.com (Erez Zilber) — Sun, 22 Nov 2009 08:51:19 UT

I think that it would be nice if we had both:
1. The average delay of a ping request.
2. A list of ping requests that almost timed out with some helpful
info (when was the ping sent and how much time until we got a
response). With this information, you can understand and debug the
whole system: you can check your target and see what caused it to be

missing references to strlcpy and strlcat when trying to compile utils/fwparam_ibft

lenn...@nutstation.com (NutStation) — Thu, 19 Nov 2009 20:05:54 UT

Hi!

I am in the process of setting up my first iscsi boot environment and
i am using open-scsi as the initiator for my nodes. I am followiing
this excellent tute: [link]

As the tute states, the fwparam_ibft executable is not part of debian
install, and so I set forth to compile it manually per the

Re: Infortrend + "iSCSI: detected conn error (1011)" + "TCP Dup ACK"

matt-openis...@alpha345.com (Matthew Dickinson) — Thu, 19 Nov 2009 20:32:48 UT

Apologies, I've attached it this time :-)

Matthew

Re: Infortrend + "iSCSI: detected conn error (1011)" + "TCP Dup ACK"

micha...@cs.wisc.edu (Mike Christie) — Thu, 19 Nov 2009 20:07:55 UT

I did not get the attachement.

If you are using Red Hat RPMs, make a red hat bugzilla
[link]. CC mchri...@redhat.com on the bugzilla or
email me at that address when you have made the bugzilla. I will then
add some network people to it. Attach your trace to the bugzilla.

Re: help iscsi-initiator discovery

micha...@cs.wisc.edu (Mike Christie) — Thu, 19 Nov 2009 20:04:32 UT

What distro are you using? If you look in /dev/disk/by-id there should
be names that are persistent across reboots.

Re: Information about iSCSI pings that almost timed out

micha...@cs.wisc.edu (Mike Christie) — Thu, 19 Nov 2009 20:02:45 UT

I think that is what I wanted. I think Erez wants something a little
different, right Erez?

Re: iscsi diagnosis help

guru.anbalag...@oracle.com (Guru Anbalagane) — Thu, 19 Nov 2009 00:50:03 UT

Hi Mike,

Thanks.
Can you please try --target=i686.

On the patches, yes, I will include it in our next VM kernel.
regards
Guru

help iscsi-initiator discovery

wwarb...@gmail.com (xi qihao) — Thu, 19 Nov 2009 08:36:53 UT

Hi all:

I a newbie of iscsi. Now I use two PC to test software iscsi.
My problem is how to fix connect of iscsi target?

The below are the detail:
SAN: iscsi target device ip: 192.168.3.2
master: iscsi initiator ip: 192.168.3.124

1. configure on SAN
1)install iscsi-target
[root@SAN ~]# yum -y install scsi-target-utils

Re: Infortrend + "iSCSI: detected conn error (1011)" + "TCP Dup ACK"

matt-openis...@alpha345.com (Matthew Dickinson) — Thu, 19 Nov 2009 15:50:29 UT

iscsi-initiator-utils-6.2.0.87 1-0.10.el5
kernel-2.6.18-164.2.1.el5

RedHat RPMs

I've tried setting this, but it appears to have no effect - it was 60, and I
increased to 360.

I've tried setting to both 16 and 32, but it behaves about the same.

I really think I'm back to the duplicate ACK problem - see the attached

Re: Information about iSCSI pings that almost timed out

ulrich.wi...@rz.uni-regensburg.de (Ulrich Windl) — Thu, 19 Nov 2009 10:10:09 UT

I thought you want to tune the timeouts. So if properly tuned, the kernel will log
when when your measurements are unusual (i.e. timeout exceeded).

So what yyou really want to have (it seems) is not just a histogram of timeouts,
but a history over time as well.

Ulrich

Re: Information about iSCSI pings that almost timed out

erezzi.l...@gmail.com (Erez Zilber) — Thu, 19 Nov 2009 09:07:38 UT

On Thu, Nov 19, 2009 at 9:38 AM, Ulrich Windl

It's in userspace, so (almost) everything is possible. It's nice to
have counters, average delay etc, but I want to be able to know
exactly when bad things almost happened (i.e. timeout almost expired).
Counters/average delay will not help me.

Erez

Re: Information about iSCSI pings that almost timed out

ulrich.wi...@rz.uni-regensburg.de (Ulrich Windl) — Thu, 19 Nov 2009 07:38:51 UT

Hi!

Wouldn't it be more obvious to calculate the average delay to a ping request?
(Possibly exponential average as for the system loads) (min and Max would be good
as well, but standard deviation probably requires use of the FPU, so that's not
possible in kernel modules (AFAIK)).

Regards,

Re: Information about iSCSI pings that almost timed out

erezzi.l...@gmail.com (Erez Zilber) — Thu, 19 Nov 2009 07:36:10 UT

The following info:
1. The exact time when the nop-out was sent.
2. How much time until we got the nop-in.

This is problematic because:
1. You can set your ping timeout to X. According to your suggestion,
we will need to have the following counters:
a. < 0.2*X
b. 0.2*X - 0.4*X
c. 0.4*X - 0.6*X

Re: iscsi diagnosis help

micha...@cs.wisc.edu (Mike Christie) — Thu, 19 Nov 2009 01:15:41 UT

Thanks Guru.

Joseph, attached is a patch that has a couple fixes from upstream in
that code path. If it does not fix your issue then I will have to send a
patch with some more debugging in it (the net layer is not waking us up
when send space is opening or it really is taking a long time to send

Re: Information about iSCSI pings that almost timed out

micha...@cs.wisc.edu (Mike Christie) — Thu, 19 Nov 2009 00:49:38 UT

What info did you want to store?

What about some perf counters for it? Add a counter for if a ping took
a-b secs, b-c secs, and c-d secs. Userspace could then just get that
info with the other stats. Also maybe a new iscsi_connection sysfs entry
that exports the avg time to complete a nop might be nice. It might help