Confirm that bnx2i iSCSI offload works for RHEL 5.5 on DELL EqualLogic
604 views
Skip to first unread message
Michael Wolf
Aug 24, 2010, 11:13:11 AM8/24/10
to open-iscsi
Hi ALL,
I try to set up oracle cluster with four nodes (DELL R610) and
EqualLogig PS6000.
Servers has four Broadcom 5709C dual network cards, two intenal, two
PCI cards.
I`m trying to get the bnx2i driver to work using iSCSI offload for
four channels.
I can't get it to work, getting error
iscsiadm: Could not login to [iface: bnx2i.00:22:19:68:3a:04, target:
iqn.2001-05.com.equallogic:0-8a0906-76c630e04-e0492d35ddc63f-a1test,
portal: 10.10.77.10,3260):
iscsiadm: initiator reported error (4 - encountered connection
failure)
Showing nothing in /var/log/messages and no events on the EQ box.
DELL suport team is not shure if bnx2i driver will work with RHEL 5.5,
they ask me to reinstall RHEL 5.2, from my opinion is that not an
option.
I am using RHEL 5.5, oracle linux, kernel 2.6.18-194.11.1.0.1.el5 on
x86_64
iscsi-initiator-utils-6.2.0.871-0.16.el5
All firmware of the R610 is up to date
All interfaces are shown as bnx2i
[root@ora-db-rac1-04 ~]# scsiadm -m iface
-bash: scsiadm: command not found
[root@ora-db-rac1-04 ~]# iscsiadm -m iface
default tcp,<empty>,<empty>,<empty>,<empty>
iser iser,<empty>,<empty>,<empty>,<empty>
bnx2i.00:22:19:68:3a:04 bnx2i,00:22:19:68:3a:
04,10.10.77.69,<empty>,<empty>
bnx2i.00:22:19:68:3a:0a bnx2i,00:22:19:68:3a:0a,
10.10.77.68,<empty>,<empty>
bnx2i.00:00:00:00:00:00 bnx2i,
00:00:00:00:00:00,<empty>,<empty>,<empty>
bnx2i.00:22:19:68:3a:06 bnx2i,00:22:19:68:3a:
06,10.10.77.67,<empty>,<empty>
bnx2i.00:10:18:64:2e:5b bnx2i,00:10:18:64:2e:5b,
10.10.77.66,<empty>,<empty>
bnx2i.00:10:18:64:98:bb bnx2i,00:10:18:64:98:bb,
10.10.77.65,<empty>,<empty>
bnx2i.00:22:19:68:3a:08 bnx2i,00:22:19:68:3a:
08,10.10.77.64,<empty>,<empty>
bnx2i.00:10:18:64:98:b9 bnx2i,
00:10:18:64:98:b9,10.10.77.63,<empty>,<empty>
Only four of them are connected to the SAN. How to figure out which hw-
adress belongs to which interface?
Is bnx2i driver supported for RHEL5.5?
Is there a more detailed description for setting up such an
environment as the RedHat manuals?
I am searching for the best way to figure out he reason for the
connect failures?
Thanks for any helping hand.
I try to set up oracle cluster with four nodes (DELL R610) and
EqualLogig PS6000.
Servers has four Broadcom 5709C dual network cards, two intenal, two
PCI cards.
I`m trying to get the bnx2i driver to work using iSCSI offload for
four channels.
I can't get it to work, getting error
iscsiadm: Could not login to [iface: bnx2i.00:22:19:68:3a:04, target:
iqn.2001-05.com.equallogic:0-8a0906-76c630e04-e0492d35ddc63f-a1test,
portal: 10.10.77.10,3260):
iscsiadm: initiator reported error (4 - encountered connection
failure)
Showing nothing in /var/log/messages and no events on the EQ box.
DELL suport team is not shure if bnx2i driver will work with RHEL 5.5,
they ask me to reinstall RHEL 5.2, from my opinion is that not an
option.
I am using RHEL 5.5, oracle linux, kernel 2.6.18-194.11.1.0.1.el5 on
x86_64
iscsi-initiator-utils-6.2.0.871-0.16.el5
All firmware of the R610 is up to date
All interfaces are shown as bnx2i
[root@ora-db-rac1-04 ~]# scsiadm -m iface
-bash: scsiadm: command not found
[root@ora-db-rac1-04 ~]# iscsiadm -m iface
default tcp,<empty>,<empty>,<empty>,<empty>
iser iser,<empty>,<empty>,<empty>,<empty>
bnx2i.00:22:19:68:3a:04 bnx2i,00:22:19:68:3a:
04,10.10.77.69,<empty>,<empty>
bnx2i.00:22:19:68:3a:0a bnx2i,00:22:19:68:3a:0a,
10.10.77.68,<empty>,<empty>
bnx2i.00:00:00:00:00:00 bnx2i,
00:00:00:00:00:00,<empty>,<empty>,<empty>
bnx2i.00:22:19:68:3a:06 bnx2i,00:22:19:68:3a:
06,10.10.77.67,<empty>,<empty>
bnx2i.00:10:18:64:2e:5b bnx2i,00:10:18:64:2e:5b,
10.10.77.66,<empty>,<empty>
bnx2i.00:10:18:64:98:bb bnx2i,00:10:18:64:98:bb,
10.10.77.65,<empty>,<empty>
bnx2i.00:22:19:68:3a:08 bnx2i,00:22:19:68:3a:
08,10.10.77.64,<empty>,<empty>
bnx2i.00:10:18:64:98:b9 bnx2i,
00:10:18:64:98:b9,10.10.77.63,<empty>,<empty>
Only four of them are connected to the SAN. How to figure out which hw-
adress belongs to which interface?
Is bnx2i driver supported for RHEL5.5?
Is there a more detailed description for setting up such an
environment as the RedHat manuals?
I am searching for the best way to figure out he reason for the
connect failures?
Thanks for any helping hand.
Mike Christie
Aug 24, 2010, 8:14:00 PM8/24/10
to open-...@googlegroups.com, Michael Wolf
On 08/24/2010 10:13 AM, Michael Wolf wrote:
> Hi ALL,
>
> I try to set up oracle cluster with four nodes (DELL R610) and
> EqualLogig PS6000.
> Servers has four Broadcom 5709C dual network cards, two intenal, two
> PCI cards.
> I`m trying to get the bnx2i driver to work using iSCSI offload for
> four channels.
> I can't get it to work, getting error
> iscsiadm: Could not login to [iface: bnx2i.00:22:19:68:3a:04, target:
> iqn.2001-05.com.equallogic:0-8a0906-76c630e04-e0492d35ddc63f-a1test,
> portal: 10.10.77.10,3260):
> iscsiadm: initiator reported error (4 - encountered connection
> failure)
>
> Showing nothing in /var/log/messages and no events on the EQ box.
> Hi ALL,
>
> I try to set up oracle cluster with four nodes (DELL R610) and
> EqualLogig PS6000.
> Servers has four Broadcom 5709C dual network cards, two intenal, two
> PCI cards.
> I`m trying to get the bnx2i driver to work using iSCSI offload for
> four channels.
> I can't get it to work, getting error
> iscsiadm: Could not login to [iface: bnx2i.00:22:19:68:3a:04, target:
> iqn.2001-05.com.equallogic:0-8a0906-76c630e04-e0492d35ddc63f-a1test,
> portal: 10.10.77.10,3260):
> iscsiadm: initiator reported error (4 - encountered connection
> failure)
>
> Showing nothing in /var/log/messages and no events on the EQ box.
So there is no iferror error in messages or nothing from iscsid?
> DELL suport team is not shure if bnx2i driver will work with RHEL 5.5,
> they ask me to reinstall RHEL 5.2, from my opinion is that not an
> option.
> I am using RHEL 5.5, oracle linux, kernel 2.6.18-194.11.1.0.1.el5 on
Is that a oracle linux kernel or the rhel 5.5 one? Are you running rhel
5.5 in a oracle linux vm?
> x86_64
> iscsi-initiator-utils-6.2.0.871-0.16.el5
Also you need the updated iscsi utils.
http://people.redhat.com/mchristi/iscsi/rhel5.6/iscsi-initiator-utils/
> All firmware of the R610 is up to date
>
> All interfaces are shown as bnx2i
> [root@ora-db-rac1-04 ~]# scsiadm -m iface
> -bash: scsiadm: command not found
> [root@ora-db-rac1-04 ~]# iscsiadm -m iface
> default tcp,<empty>,<empty>,<empty>,<empty>
> iser iser,<empty>,<empty>,<empty>,<empty>
> bnx2i.00:22:19:68:3a:04 bnx2i,00:22:19:68:3a:
> 04,10.10.77.69,<empty>,<empty>
Are these IPs on the same subnet as the network port they are attached to?
> bnx2i.00:22:19:68:3a:0a bnx2i,00:22:19:68:3a:0a,
> 10.10.77.68,<empty>,<empty>
> bnx2i.00:00:00:00:00:00 bnx2i,
> 00:00:00:00:00:00,<empty>,<empty>,<empty>
> bnx2i.00:22:19:68:3a:06 bnx2i,00:22:19:68:3a:
> 06,10.10.77.67,<empty>,<empty>
> bnx2i.00:10:18:64:2e:5b bnx2i,00:10:18:64:2e:5b,
> 10.10.77.66,<empty>,<empty>
> bnx2i.00:10:18:64:98:bb bnx2i,00:10:18:64:98:bb,
> 10.10.77.65,<empty>,<empty>
> bnx2i.00:22:19:68:3a:08 bnx2i,00:22:19:68:3a:
> 08,10.10.77.64,<empty>,<empty>
> bnx2i.00:10:18:64:98:b9 bnx2i,
> 00:10:18:64:98:b9,10.10.77.63,<empty>,<empty>
>
> Only four of them are connected to the SAN. How to figure out which hw-
> adress belongs to which interface?
Start the iscsi service (or at least make sure bnx2i is loaded) then run:
[root@meanminna kernel]# iscsiadm -m host -P 1
Host Number: 8
State: running
Transport: bnx2i
Initiatorname: <empty>
IPaddress: <empty>
HWaddress: 00:10:18:15:fd:f1
Netdev: eth1
You can see the network device the iscsi offload interface is attached
to (in this case eth1), and you can see the MAC/hwaddress for the iscsi
interface.
> Is bnx2i driver supported for RHEL5.5?
Yes.
> Is there a more detailed description for setting up such an
> environment as the RedHat manuals?
Section 5.1.2 Setting up a iface for a iSCSI offload card of the iscsi
readme in /usr/share/docs/iscsi-initiator-utils-$VERSION/README
Michael Wolf
Aug 25, 2010, 3:55:50 AM8/25/10
to open-iscsi
>
> - Zitierten Text anzeigen -- Zitierten Text ausblenden -
>
> - Zitierten Text anzeigen -
Hi Mike,
Trying to connect to the EQL box there are no messagese at all.
On these oracle related boxes we use the oracle linux kernel. We do
not use oracle linux vm.
Are there any issues with this kernel and iscsi?
I have downloaded the update and will install and test with this
release, I will come back to you with the results.
All storage IPs are on the same subnet 10.10.77.00. The EQL box
listens to 10.10.77.10.
Kind regards
Michael
Michael Wolf
Aug 25, 2010, 6:10:23 AM8/25/10
to open-iscsi
>
> - Zitierten Text anzeigen -
With the new iscsi-initiator-utils set up, I now can connect to the
> - Zitierten Text anzeigen -
target.
I have set up two bnx2i ports with IP`s 63 and 64:
[root@ora-db-rac1-04 ~]# iscsiadm -m iface
default tcp,<empty>,<empty>,<empty>,<empty>
iser iser,<empty>,<empty>,<empty>,<empty>
bnx2i.00:22:19:68:3a:04 bnx2i,00:22:19:68:3a:
bnx2i.00:22:19:68:3a:0a bnx2i,00:22:19:68:3a:0a,
10.10.77.64,<empty>,<empty>
bnx2i.00:00:00:00:00:00 bnx2i,
00:00:00:00:00:00,<empty>,<empty>,<empty>
bnx2i.00:22:19:68:3a:06 bnx2i,00:22:19:68:3a:
06,<empty>,<empty>,<empty>
00:00:00:00:00:00,<empty>,<empty>,<empty>
bnx2i.00:22:19:68:3a:06 bnx2i,00:22:19:68:3a:
bnx2i.00:10:18:64:2e:5b bnx2i,00:10:18:64:2e:
5b,<empty>,<empty>,<empty>
bnx2i.00:10:18:64:98:bb bnx2i,
00:10:18:64:98:bb,<empty>,<empty>,<empty>
bnx2i.00:22:19:68:3a:08 bnx2i,00:22:19:68:3a:
08,10.10.77.63,<empty>,<empty>
bnx2i.00:10:18:64:98:b9 bnx2i,
00:10:18:64:98:b9,<empty>,<empty>,<empty>
I have done a discovery, finding two targets, my test target aitest
and the vss-Control target from the EQL box.
[root@ora-db-rac1-04 ~]# iscsiadm -m discovery -t st -p
10.10.77.10:3260 -I bnx2i.00:22:19:68:3a:08 -P 1
Target: iqn.2001-05.com.equallogic:0-8a0906-68d84cc03-cbc00013c9149355-
vss-control
Portal: 10.10.77.10:3260,1
Iface Name: bnx2i.00:22:19:68:3a:08
Target: iqn.2001-05.com.equallogic:0-8a0906-76c630e04-e0492d35ddc4c63f-
a1test
Portal: 10.10.77.10:3260,1
Iface Name: bnx2i.00:22:19:68:3a:08
Now connecting to the test target:
[root@ora-db-rac1-04 ~]# iscsiadm -m node -T iqn.
2001-05.com.equallogic:0-8a0906-76c630e04-e0492d35ddc4c63f-a1test -p
10.10.77.10:3260 -I bnx2i.00:22:19:68:3a:08 -l
Logging in to [iface: bnx2i.00:22:19:68:3a:08, target: iqn.
2001-05.com.equallogic:0-8a0906-76c630e04-e0492d35ddc4c63f-a1test,
portal: 10.10.77.10,3260]
Login to [iface: bnx2i.00:22:19:68:3a:08, target: iqn.
2001-05.com.equallogic:0-8a0906-76c630e04-e0492d35ddc4c63f-a1test,
portal: 10.10.77.10,3260]: successful
Everything is lookingfine, but there are a lot of failures in /var/log/
messages:
Aug 25 12:03:56 ora-db-rac1-04 kernel: bnx2i [02:00.00]: ISCSI_INIT
passed
Aug 25 12:03:57 ora-db-rac1-04 kernel: connection4:0: detected conn
error (1011)
Aug 25 12:03:57 ora-db-rac1-04 kernel: bnx2i [02:00.00]: ISCSI_INIT
passed
Aug 25 12:03:58 ora-db-rac1-04 kernel: connection4:0: bnx2i: conn
update - MBL 0x40000 FBL 0x10000MRDSL_I 0x40000 MRDSL_T 0x10000
Aug 25 12:03:58 ora-db-rac1-04 kernel: Vendor: EQLOGIC Model:
100E-00 Rev: 4.2
Aug 25 12:03:58 ora-db-rac1-04 kernel: Type: Direct-
Access ANSI SCSI revision: 05
Aug 25 12:03:58 ora-db-rac1-04 kernel: SCSI device sdb: 215040 512-
byte hdwr sectors (110 MB)
Aug 25 12:03:58 ora-db-rac1-04 kernel: sdb: Write Protect is off
Aug 25 12:03:58 ora-db-rac1-04 kernel: SCSI device sdb: drive cache:
write through
Aug 25 12:03:58 ora-db-rac1-04 kernel: SCSI device sdb: 215040 512-
byte hdwr sectors (110 MB)
Aug 25 12:03:58 ora-db-rac1-04 kernel: sdb: Write Protect is off
Aug 25 12:03:58 ora-db-rac1-04 kernel: SCSI device sdb: drive cache:
write through
Aug 25 12:03:58 ora-db-rac1-04 kernel: sdb: unknown partition table
Aug 25 12:03:58 ora-db-rac1-04 kernel: sd 8:0:0:0: Attached scsi disk
sdb
Aug 25 12:03:58 ora-db-rac1-04 multipathd: sdb: add path (uevent)
Aug 25 12:03:58 ora-db-rac1-04 kernel: sd 8:0:0:0: Attached scsi
generic sg3 type 0
Aug 25 12:03:58 ora-db-rac1-04 iscsid: connection4:0 is operational
now
Aug 25 12:03:58 ora-db-rac1-04 iscsid: Could not write to /sys/bus/
scsi/devices/8:0:0:0/queue_depth. Invalid permissions.
Aug 25 12:03:58 ora-db-rac1-04 iscsid: Could not queue depth for LUN 0
err 13.
Aug 25 12:04:58 ora-db-rac1-04 multipathd: sdb: failed to get path uid
Aug 25 12:04:58 ora-db-rac1-04 multipathd: uevent trigger error
Aug 25 12:05:11 ora-db-rac1-04 kernel: connection4:0: detected conn
error (1011)
Aug 25 12:05:11 ora-db-rac1-04 iscsid: Kernel reported iSCSI
connection 4:0 error (1011) state (3)
Aug 25 12:05:13 ora-db-rac1-04 kernel: connection4:0: detected conn
error (1011)
Aug 25 12:05:14 ora-db-rac1-04 kernel: bnx2i [02:00.00]: ISCSI_INIT
passed
Aug 25 12:05:14 ora-db-rac1-04 kernel: connection4:0: detected conn
error (1011)
Aug 25 12:05:14 ora-db-rac1-04 iscsid: Kernel reported iSCSI
connection 4:0 error (1011) state (1)
Aug 25 12:05:14 ora-db-rac1-04 kernel: bnx2i [02:00.00]: ISCSI_INIT
passed
Aug 25 12:05:15 ora-db-rac1-04 kernel: connection4:0: bnx2i: conn
update - MBL 0x40000 FBL 0x10000MRDSL_I 0x40000 MRDSL_T 0x10000
Aug 25 12:05:15 ora-db-rac1-04 kernel: session4: target reset
succeeded
Aug 25 12:05:15 ora-db-rac1-04 iscsid: connection4:0 is operational
after recovery (1 attempts)
Aug 25 12:06:26 ora-db-rac1-04 kernel: connection4:0: detected conn
error (1011)
Aug 25 12:06:27 ora-db-rac1-04 iscsid: Kernel reported iSCSI
connection 4:0 error (1011) state (3)
Aug 25 12:06:29 ora-db-rac1-04 kernel: bnx2i [02:00.00]: ISCSI_INIT
passed
Aug 25 12:06:30 ora-db-rac1-04 kernel: connection4:0: detected conn
error (1011)
Aug 25 12:06:30 ora-db-rac1-04 kernel: bnx2i [02:00.00]: ISCSI_INIT
passed
Aug 25 12:06:31 ora-db-rac1-04 kernel: connection4:0: bnx2i: conn
update - MBL 0x40000 FBL 0x10000MRDSL_I 0x40000 MRDSL_T 0x10000
Aug 25 12:06:31 ora-db-rac1-04 iscsid: connection4:0 is operational
after recovery (1 attempts)
Aug 25 12:06:51 ora-db-rac1-04 kernel: connection4:0: detected conn
error (1011)
Aug 25 12:06:51 ora-db-rac1-04 iscsid: Kernel reported iSCSI
connection 4:0 error (1011) state (3)
Aug 25 12:06:53 ora-db-rac1-04 kernel: bnx2i [02:00.00]: ISCSI_INIT
passed
Aug 25 12:06:53 ora-db-rac1-04 kernel: connection4:0: detected conn
error (1011)
Aug 25 12:06:54 ora-db-rac1-04 kernel: bnx2i [02:00.00]: ISCSI_INIT
passed
Aug 25 12:06:54 ora-db-rac1-04 kernel: connection4:0: bnx2i: conn
update - MBL 0x40000 FBL 0x10000MRDSL_I 0x40000 MRDSL_T 0x10000
Aug 25 12:06:54 ora-db-rac1-04 kernel: session4: target reset
succeeded
Aug 25 12:06:55 ora-db-rac1-04 iscsid: connection4:0 is operational
after recovery (1 attempts)
Aug 25 12:07:49 ora-db-rac1-04 kernel: INFO: task vol_id:5911 blocked
for more than 120 seconds.
Aug 25 12:07:49 ora-db-rac1-04 kernel: "echo 0 > /proc/sys/kernel/
hung_task_timeout_secs" disables this message.
Aug 25 12:07:49 ora-db-rac1-04 kernel: vol_id D
ffffffff80150989 0 5911 1 5523 (NOTLB)
Aug 25 12:07:49 ora-db-rac1-04 kernel: ffff810614ef7c08
0000000000000082 ffff8106295c4e00 0000000000000018
Aug 25 12:07:49 ora-db-rac1-04 kernel: 0000000000000056
0000000000000006 ffff810619f5c820 ffffffff80309b60
Aug 25 12:07:49 ora-db-rac1-04 kernel: 0000033c75c6b1de
00000000000089a1 ffff810619f5ca08 0000000000000000
Aug 25 12:07:49 ora-db-rac1-04 kernel: Call Trace:
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff8006e1f7>]
do_gettimeofday+0x40/0x90
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff800289e0>] sync_page
+0x0/0x43
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff800637ea>]
io_schedule+0x3f/0x67
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff80028a1e>] sync_page
+0x3e/0x43
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff8006392e>]
__wait_on_bit_lock+0x36/0x66
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff8003fbc5>]
__lock_page+0x5e/0x64
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff800a0a5d>]
wake_bit_function+0x0/0x23
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff8000c2e4>]
do_generic_mapping_read+0x1df/0x359
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff8000d0fd>]
file_read_actor+0x0/0x159
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff8000c5aa>]
__generic_file_aio_read+0x14c/0x198
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff800c67cb>]
generic_file_read+0xac/0xc5
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff800a0a2f>]
autoremove_wake_function+0x0/0x2e
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff800b765c>]
audit_syscall_entry+0x180/0x1b3
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff8000b69a>] vfs_read
+0xcb/0x171
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff80011bb3>] sys_read
+0x45/0x6e
Aug 25 12:07:49 ora-db-rac1-04 kernel: [<ffffffff8005d28d>] tracesys
+0xd5/0xe0
Aug 25 12:07:49 ora-db-rac1-04 kernel:
[root@ora-db-rac1-04 ~]#
Any idea?
Regards Michael
Ciprian Marius Vizitiu (GBIF)
Aug 25, 2010, 6:43:05 AM8/25/10
to open-...@googlegroups.com
Michael Wolf wrote:
> Only four of them are connected to the SAN. How to figure out which hw-
> adress belongs to which interface?
>
Perhaps BIOS / iface's ROM util should give you some hint(s).> Only four of them are connected to the SAN. How to figure out which hw-
> adress belongs to which interface?
>
Shyam Iyer
Aug 25, 2010, 1:50:47 PM8/25/10
to open-...@googlegroups.com
Mike Christie
Aug 25, 2010, 2:35:04 PM8/25/10
to open-...@googlegroups.com, Michael Wolf
On 08/25/2010 02:55 AM, Michael Wolf wrote:
> On these oracle related boxes we use the oracle linux kernel. We do
> not use oracle linux vm.
> Are there any issues with this kernel and iscsi?
>
> On these oracle related boxes we use the oracle linux kernel. We do
> not use oracle linux vm.
> Are there any issues with this kernel and iscsi?
>
What do you mean you use RHEL 5.5 then? Are you using RHEL 5.5 as a base
but then replace that kernel with the oracle linux one? Why do you do
that btw?
Mike Christie
Aug 25, 2010, 3:00:14 PM8/25/10
to open-...@googlegroups.com, Michael Wolf, Eddie Wai
Adding Eddie from Broadcom.
If the oracle linux kernel is based on rhel 5.5's then I think to work
with equallogic boxes you might need to have the target not send nops as
pings, because the rhel 5.5 bnx2i driver had some bugs in that code.
Eddie, was there anything else?
Ulrich Windl
Aug 26, 2010, 4:46:54 AM8/26/10
to open-...@googlegroups.com
>>> Shyam Iyer <shyam...@dell.com> schrieb am 25.08.2010 um 19:50 in Nachricht
<4C7557F7...@dell.com>:
<4C7557F7...@dell.com>:
For a HP DL380G5 the ILO2 reports these under "System Information"/"NIC":
Port 1 NIC MAC address: 00:1A:4B:4D:25:B2
Port 2 NIC MAC address: 00:1A:4B:4D:25:B0
iLO 2: 00:1A:4B:4D:25:B4
Port 1 iSCSI MAC address: 00:1A:4B:4D:25:B3
Port 2 iSCSI MAC address: 00:1A:4B:4D:25:B1
So maybe it's a good idea to look at your management interface...
Ulrich
Michael Wolf
Aug 26, 2010, 10:23:31 AM8/26/10
to open-iscsi
On 26 Aug., 10:46, "Ulrich Windl" <Ulrich.Wi...@rz.uni-regensburg.de>
wrote:
> >>> Shyam Iyer <shyam_i...@dell.com> schrieb am 25.08.2010 um 19:50 in Nachricht
>
> <4C7557F7.9000...@dell.com>:
>
> > On 08/25/2010 06:43 AM, Ciprian Marius Vizitiu (GBIF) wrote:
> > > Michael Wolf wrote:
> > >> Only four of them are connected to the SAN. How to figure out which hw-
> > >> adress belongs to which interface?
> > > Perhaps BIOS / iface's ROM util should give you some hint(s).
>
> > Offload HW address = ethX HW Address + 1
>
> For a HP DL380G5 the ILO2 reports these under "System Information"/"NIC":
> Port 1 NIC MAC address: 00:1A:4B:4D:25:B2
> Port 2 NIC MAC address: 00:1A:4B:4D:25:B0
> iLO 2: 00:1A:4B:4D:25:B4
> Port 1 iSCSI MAC address: 00:1A:4B:4D:25:B3
> Port 2 iSCSI MAC address: 00:1A:4B:4D:25:B1
>
> So maybe it's a good idea to look at your management interface...
>
> Ulrich
Hi All,
> > On 08/25/2010 06:43 AM, Ciprian Marius Vizitiu (GBIF) wrote:
> > > Michael Wolf wrote:
> > >> Only four of them are connected to the SAN. How to figure out which hw-
> > >> adress belongs to which interface?
> > > Perhaps BIOS / iface's ROM util should give you some hint(s).
>
> > Offload HW address = ethX HW Address + 1
>
> For a HP DL380G5 the ILO2 reports these under "System Information"/"NIC":
> Port 1 NIC MAC address: 00:1A:4B:4D:25:B2
> Port 2 NIC MAC address: 00:1A:4B:4D:25:B0
> iLO 2: 00:1A:4B:4D:25:B4
> Port 1 iSCSI MAC address: 00:1A:4B:4D:25:B3
> Port 2 iSCSI MAC address: 00:1A:4B:4D:25:B1
>
> So maybe it's a good idea to look at your management interface...
>
> Ulrich
Thank you for the information on hw addresses.
It`s clear now.
We use OEL 5.5 kernel which is based on RHEL 5.5.
Is it correct to set the parameters
node.conn[0].timeo.noop_out_timeout = 0
node.conn[0].timeo.noop_out_interval = 0
to work with equallogic boxes?
node.session.iscsi.FastAbort = No
Any other issues with config files?
On a second box I had done a fresh RHEL 5.5 installation from scratch.
When I have problems here I think the best way is to open a new
discussion for that.
Hopefully original RHEL will work better than OEL.
Mike Christie
Aug 26, 2010, 4:34:19 PM8/26/10
to open-...@googlegroups.com, Michael Wolf
This will tell the initiator not to send nops as pings. But the target
will still send its own pings by default. You need to contact equalogic
to see how to turn them off.
Michael Wolf
Sep 10, 2010, 9:10:18 AM9/10/10
to open-iscsi
>
> - Zitierten Text anzeigen -
Hi all,
> - Zitierten Text anzeigen -
I think I was able to find a solution for the connectivity failures.
1) For the connectivity beetween hosts and EQ box I use DELL 6248
switches. The hosts interfacese, switches and EQ box are confiured to
use jumbo frames (MTU=9000). The reason for my connectivity problems
was, that the DELL switches need to be configured with Maximum Frame
Size = 9216. 9000 is not enougth. After changing this parameter it
works. The other option is to use standard frame size on all devices
MTU 1500.
2) The Broadcom adapters and there aktual driver does not support
multipathing. I was able to connect with one adapter whithout problem,
but I was not able to get an second connection to the EQ-Box.
Thanks for your help.
Michael
Reply all
Reply to author
Forward
0 new messages