What does 1 mean? Surely not 1 byte MTU :)
IMO a good format would be the MTU value in bytes.
E.g. 512, 1024, 2048, 4096.
Quoting Shirley Ma [EMAIL PROTECTED]:
Subject: RE: Re: openSM: Different IB MTUs
Eitan,
That's a good approach to address the issue.
thanks
Shirley Ma
Inactive hide
Quoting Amar Mudrankit [EMAIL PROTECTED]:
Subject: ARP in IPoIB
Hello all,
Being new to this group, following questions may sound a bit basic level
but I would be really very happy if somebody could help me out in those.
1] Does current implementation of IPoIB supports IP hosts
Eliminate double-kfree in mlx4_mr_alloc error flow.
Signed-off-by: Jack Morgenstein [EMAIL PROTECTED]
Index: connectx/drivers/net/mlx4/mr.c
===
--- connectx.orig/drivers/net/mlx4/mr.c 2007-07-26 10:04:57.0 +0300
+++
scsi/scsi_transport_iscsi.h uses struct mutex and struct list_head,
so while linux/mutex.h and linux/list.h seem to be pulled in indirectly
by one of the headers it includes, the right thing
is to include linux/mutex.h and linus/list.h directly.
Signed-off-by: Michael S. Tsirkin [EMAIL PROTECTED]
Hi,
ibv_modify_qp() fails with return value 22 when I try to open a new CM
connection under load (already ~3000 RDMA connections opened). I tried
to figure out what return value 22 means but could not find it in the
mthca kernel driver.
Any hints? What does return value 22 mean?
I use OFED-1.1
Hi.
Joerg Zinke wrote:
Hi,
ibv_modify_qp() fails with return value 22 when I try to open a new CM
connection under load (already ~3000 RDMA connections opened). I tried
to figure out what return value 22 means but could not find it in the
mthca kernel driver.
Any hints? What does return value
Sean Hefty wrote:
This file has *all* the possible keywords.
The administrator really doesn't have to use them all.
For instance, there are three different ways to define port groups:
- by guid list
- by node type
- by port names
You could stick with the guids only - this gives you all the
This email was generated automatically, please do not reply
git_url: git://git.openfabrics.org/ofed_1_2/linux-2.6.git
git_branch: ofed_1_2
Common build parameters: --with-ipoib-mod --with-sdp-mod --with-srp-mod
--with-user_mad-mod --with-user_access-mod --with-mthca-mod --with-core-mod
On Thu, Jul 26, 2007 at 07:58:18AM +0300, Erez Zilber wrote:
[...]
I guess that you meant that you're using only the UNH iSCSI target
(because the iSER initiator should be used with open-iscsi).
Well, we are actually using both initiator and target. We grabed an
older version of the
But again, the administrator doesn't *have* to use all these.
He can simply define sl2vl-tables, and then match service-id
(in qos-match-rules) to a certain sl (in qos-levels).
That's it.
No MTU, rate, packet lifetime or any other low level data.
Does the following file look better?
My take is
In the socket API, socket options describe what protocol they are
intended for. You can have options that are intended for IP or TCP and
other protocol layers.
We could do some rdma_setopt() interface, and define both transport
independent options and transport-specific options. Then if
Thanks for the suggestion Sasha!
Our host stack does receive a rereregistration notice and does resubscribe
all handlers at
that point in time. At the time of the SM migration, our stack prints out
some informational messages to
confirm this:
Jul 18 14:31:09 localhost kernel: Event
This same argument is why IPoIB ARP decisions really have no bearing
on IB PRs. IPoIB ARP logic and refreshes is designed to support the
distributed ND lookup model - IB PR's have completely different
lifetime rules that are totally unrelated to ARP's liftime rules. The
existing trap monitoring
Do not fetch initial switch LFTs for discovered switches. OpenSM doesn't
use it anyway, but it creates additional subnet traffic.
Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED]
---
opensm/opensm/osm_sw_info_rcv.c |6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git
This removes static variable __some_hop_count_set from osm_ucast_mgr
and instead uses flag stored as structure memmber.
Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED]
---
opensm/include/opensm/osm_ucast_mgr.h |6 ++
opensm/opensm/osm_ucast_mgr.c | 16
2
As was discussed previously on the list this moves ucast and mcast
dumper functions to separate file (osm_dump.c). Dump generators will
be invoked after heavy sweep.
Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED]
---
opensm/include/opensm/osm_opensm.h |4 +
opensm/opensm/Makefile.am
Hi Eitan,
On 09:26 Thu 26 Jul , Eitan Zahavi wrote:
I am happy you actually use the simulator.
Please provide more info regarding the failure. You should tar compress
the /tmp/ibmgtsim. of your run.
I can send this for you if you want, but the failure is trivial.
6. The default
OpenSM: More changes from osm.log to opensm.log
Signed-off-by: Hal Rosenstock [EMAIL PROTECTED]
diff --git a/opensm/opensm/main.c b/opensm/opensm/main.c
index 8038dd3..f3429ff 100644
--- a/opensm/opensm/main.c
+++ b/opensm/opensm/main.c
@@ -253,7 +253,7 @@ show_usage(void)
printf( -f\n
I believe a better solution is for everyone to use cached records, if
they exist, with a feedback mechanism from the CM that removes paths on
a connection failure or path migration event.
Ack timeout on an RC QP is also a good indication we should redo the lookup.
--
MST
Quoting [EMAIL PROTECTED] [EMAIL PROTECTED]:
Subject: Re: [RFC/PATCH] mthca: ensure alignment of doorbell writes
On Thu, Jul 26, 2007 at 06:39:46AM +0300, Michael S. Tsirkin wrote:
These should be getting 'union mthca_doorbell *db' I think.
Hi Michael;
Want to make sure I
In message [EMAIL PROTECTED]
m,Shirley Ma writes:
Set default as 4 (2K) is more proper than 1(512?). All HCAs support 2K =
at
least now.
dont some devices perform better with 1k mtu's? in particular, any
device that suffers from the 'tavor quirk'.
___
On Thu, 26 Jul 2007, Michael S. Tsirkin wrote:
Now that the 2.6.23 merge window is close and people have time to
review new code, we are hoping for more comments.
You might want to send copy of patches to openfabrics general and
lkml if you do.
Good idea. We were already planning
attachment: miracle01.gif___
general mailing list
general@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
Now that the 2.6.23 merge window is close and people have time to
review new code, we are hoping for more comments.
You might want to send copy of patches to openfabrics general and lkml if you
do.
--
MST
___
general mailing list
Will you send patches for iSER soon? I'd like to test it, and make sure
that iSER over IB is not damaged.
Our patches may not interest you since we are using an older version of
the iSER code. However, we will also be exploring the use of IB with our
implementations. Will this
Hello all,
Being new to this group, following questions may sound a bit basic level
but I would be really very happy if somebody could help me out in those.
1] Does current implementation of IPoIB supports IP hosts spanning different
IB subnets? If so, how does it resolves the IP
Hi Sasha,
I am happy you actually use the simulator.
Please provide more info regarding the failure. You should tar compress
the /tmp/ibmgtsim. of your run.
The following flow is performed by this test:
1. Three partitions are created with random Pkeys. The first 2 will have
full members.
This removes actually useless subn.opt.reassign_lfts parameter.
Its value is used only for initial setup of ignore_existing_lfts flag.
But later this flag becomes unconditionally TRUE if at least one switch
is found in the fabric. If not (and fabric is just back to back
connected CAs) there is no
On 08:58 Thu 26 Jul , Hal Rosenstock wrote:
On 7/26/07, Eitan Zahavi [EMAIL PROTECTED] wrote:
*I propose that when there is no MTU in the partition policy file
OpenSM use a *
*configurable default from: **/etc/cache/opensm/opensm.opt.*
That would make this the default rather
Set default as 4 (2K) is more proper than 1(512?). All HCAs support 2K at
least now.
Thanks
Shirley Ma
Michael S.
Tsirkin
On Thu, 26 Jul 2007, Doug Ledford wrote:
On Tue, 2007-07-24 at 12:20 -0500, Tom Tucker wrote:
For those interested in NFS-RDMA, OGC has created an install package
based on the OFA 1.2 GA release. The package supports both SLES 10 and
RHEL 5. You can download this package from
Sean Hefty wrote:
Steve,
Do you have any input with respect to how the RDMA CM selects and maps
QoS (priority, traffic class, VLAN, flow label, etc.)? (See below)
Hide the QoS selection under the current interface? Use the IPv6
flowinfo field? Rely on destination port? Input QoS through
From: Amar Mudrankit
Michael,
thanks for your reply. But, this gives rise to couple of questions..
1] If such multicast routing protocol for IB routers is not yet specifid
by IBTA or IETF, then current implementation have IP subnet restricted
within a IB
On 7/26/07, Eitan Zahavi [EMAIL PROTECTED] wrote:
*I propose that when there is no MTU in the partition policy file
OpenSM use a *
*configurable default from: **/etc/cache/opensm/opensm.opt.*
That would make this the default rather than 2K. IMO it should be when some
special unused mtu is
This email was generated automatically, please do not reply
git_url: git://git.openfabrics.org/ofed_1_2/linux-2.6.git
git_branch: ofed_1_2_c
Common build parameters: --with-ipoib-mod --with-sdp-mod --with-srp-mod
--with-user_mad-mod --with-user_access-mod --with-mthca-mod --with-mlx4-mod
Hi Eitan, Hal,
On 20:44 Wed 25 Jul , Eitan Zahavi wrote:
I am not following you.
Why do a user need to run -y if a simple legal cable connector is
plugged?
Because duplicated GUIDs detector can aborts OpenSM when
regular port is reconnected to another location during hard
On 12:37 Thu 26 Jul , lbt wrote:
Thanks for the suggestion Sasha!
Our host stack does receive a rereregistration notice and does resubscribe
all handlers at
that point in time. At the time of the SM migration, our stack prints out
some informational messages to
confirm this:
On 09:25 Thu 26 Jul , Eitan Zahavi wrote:
Hi Eitan, Hal,
On 20:44 Wed 25 Jul , Eitan Zahavi wrote:
I am not following you.
Why do a user need to run -y if a simple legal cable connector is
plugged?
Because duplicated GUIDs detector can aborts OpenSM when
On 13:53 Thu 26 Jul , Hal Rosenstock wrote:
OpenSM: More changes from osm.log to opensm.log
Signed-off-by: Hal Rosenstock [EMAIL PROTECTED]
Applied. Thanks.
Sasha
___
general mailing list
general@lists.openfabrics.org
On 09:00 Thu 26 Jul , Eitan Zahavi wrote:
I propose that when there is no MTU in the partition policy file OpenSM
use a
configurable default from: /etc/cache/opensm/opensm.opt.
Something like:
# The default MTU to be used for IPoIB and other MCGs when the
partition-policy
# does not
On Tue, 2007-07-24 at 12:20 -0500, Tom Tucker wrote:
For those interested in NFS-RDMA, OGC has created an install package
based on the OFA 1.2 GA release. The package supports both SLES 10 and
RHEL 5. You can download this package from
http://www.opengridcomputing.com/nfs-rdma.html.
Please
Hi Sasha,
Sasha Khapyorsky wrote:
Hi Yevgeny,
Some initial comments.
On 01:07 Sun 22 Jul , Yevgeny Kliteynik wrote:
Hi All
Please find the attached RFC describing how QoS policy support could be
implemented in the OpenFabrics stack.
Your comments are welcome.
-- Yevgeny
Hi.
I'm trying to install OFED-1.2 on x86 (32bit) debian machine.
Although build_env.sh seems to work on debian,
it fails compiling both kernel modules and user land tools by rpmbuild.
Is OFED-1.2 tested on debian or totally unsupported?
Thanks,
Yoshi Tamura
OSM Simulation Regression Summary
[Generated mail - please do NOT reply]
OpenSM rev = Tue_Jul_24_09:41:51_2007 [feadbcd8281f007f092bbd95a2e078cac5a8a0aa]
ibutils rev = Tue_Mar_13_14:36:32_2007
[80aaff94f0eb65117db39b9db7d609ffdcc055de]
Total=560 Pass=560 Fail=0
Pass:
42 Stability
On 09:26 Thu 26 Jul , Eitan Zahavi wrote:
I am happy you actually use the simulator.
Please provide more info regarding the failure. You should tar
compress the /tmp/ibmgtsim. of your run.
I can send this for you if you want, but the failure is trivial.
No need if you
Eitan,
That's a good approach to address the issue.
thanks
Shirley Ma
Eitan Zahavi
[EMAIL PROTECTED]
Hi,
On Thu, 26 Jul 2007 13:00:01 +0300
Dotan Barak [EMAIL PROTECTED] wrote:
Joerg Zinke wrote:
ibv_modify_qp() fails with return value 22 when I try to open a new
CM connection under load (already ~3000 RDMA connections opened). I
tried to figure out what return value 22 means but could
OpenSM/osm_port.c: Fix opvls and neighbormtu when remote port invalid
Signed-off-by: Hal Rosenstock [EMAIL PROTECTED]
diff --git a/opensm/opensm/osm_port.c b/opensm/opensm/osm_port.c
index e03e316..b9c52f4 100644
--- a/opensm/opensm/osm_port.c
+++ b/opensm/opensm/osm_port.c
@@ -387,12 +388,12 @@
On Thu, Jul 26, 2007 at 06:39:46AM +0300, Michael S. Tsirkin wrote:
These should be getting 'union mthca_doorbell *db' I think.
Hi Michael;
Want to make sure I understand your point. Are you saying, e.g.,
that the function:
static inline void mthca_ring_db(union mthca_doorbell db,
On 14:16 Thu 26 Jul , Hal Rosenstock wrote:
include/iba/ib_types.h: Fix comment
Signed-off-by: Hal Rosenstock [EMAIL PROTECTED]
Applied. Thanks.
Sasha
___
general mailing list
general@lists.openfabrics.org
50 matches
Mail list logo