Problem with MSL6000 and In_sync state in switch fc 300

[i5513]

Hi,

I'm having trouble with a MSL6000 and its communication by fiber channel throwgh a switch Brocade to a HP server.

I tryied to change the wire but it still fail.

Problem:

- Until two weeks ago all was working ok
- Then it stopped to work (this message was displayed a lot of times):
Oct 23 15:31:04 backup kernel: st0: Error 70000 (sugg. bt 0x0, driver
bt 0x0, host bt 0x7).

This was triggered by a backup software operation (bacula in this case) try.

- Investing this issue:
# mtx -f /dev/autochanger1 status
cannot open SCSI device '/dev/autochanger1' - No such file or
directory
# ls -l /dev/autochanger1
lrwxrwxrwx 1 root root 3 2009-10-30 09:47 /dev/autochanger1 -> sg4
The device MSL6000 seems to be present in the system:
# lsscsi -g
...
[0:0:2:0] mediumx HP MSL6000 Series 0520 /dev/sch0 /dev/
sg4
[0:0:2:1] tape HP Ultrium 3-SCSI G63W /dev/st2 /dev/
sg5
[0:0:2:2] tape HP Ultrium 3-SCSI G63W /dev/st3 /dev/
sg6
[0:0:2:3] storage HP NS E1200-320 571f - /dev/
sg7
[1:0:3:0] tape HP Ultrium 3-SCSI G63W /dev/st0 /dev/
sg2
[1:0:4:0] tape HP Ultrium 3-SCSI G54W /dev/st1 /dev/
sg3

- I tried to remove and add the device:
# echo "scsi remove-single-device 0 0 2 0" > /proc/scsi/scsi
# echo "scsi add-single-device 0 0 2 0" > /proc/scsi/scsi
# dmesg
lpfc 0000:03:01.0: 0:0713 SCSI layer issued LUN reset (2, 0) Data:
x2002 x3 x2
lpfc 0000:03:01.0: 0:0714 SCSI layer issued Bus Reset Data: x2002
scsi 0:0:2:0: scsi: Device offlined - not ready after error recovery

I had to turn off the MSL6000 and then turn on, and then server saw again the devices and mtx works again. (Too in fc switch (zone viewer) I saw msl device without any color and after turn on it is with color again)

One hour latter, again the same issue.

Then I look at fc switch and I discovered the port where msl6000 is attached in In_Sync state many many time.

I tried changing the port ... with no succesfull result.

How can I test fc from MSL6000 ? It is possible to know if its hba (or like it is named) is working fine?

Do you know what In_sync is mean ?

Thank you very much, any help is appreciated

[i5513]

May this could be caused by licenses issues ?

My port is in In_sync when it doesn't work.

When I turn off and turn on MSL6000 it works 1 hour more or less then port state go to In_sync.

My system tells:
Nov 5 11:52:42 backup kernel: lpfc 0000:03:01.0: 0:0203 Nodev timeout on WWPN 10:0:0:e0:2:23:a7:c5 NPort x11500 Data: x2000008 x7 x6
Nov 5 11:52:47 backup kernel: rport-0:0-8: blocked FC remote port time out: removing target and saving binding

And in brocade manual switch I can read (Add Configuring Dynamic Ports on Demand Section):
If the switch detects more active links than allowed by the current POD licenses, then some ports will
not be assigned a POD license. Ports that do not receive a POD assignments have a state of No Sync or
In Sync; these ports are not allowed to progress to the online state. Ports that cannot be brought online
because insufficient POD licenses have a state of (No POD License) Disabled. (Use the switchShow
command to display the port states.)


> licenseshow
bycSQQezz9czdTRM:
Web license
Zoning license
Fabric license
RRycezeSRSSi3dfZ:
Full Ports on Demand license - additional 16 port upgrade

However:
> licenseport --show
licenseport failed : this command is not supported on this product.

It is a license issue ?

Thanks!

[Torsten.]

First I would update the NSR firmware and check the FC settings of the NSR. Point your browser to the NSR management page.

[i5513]

How can I know the nsr ip ?

I can log in msl6000 by web where I can see the next menus:

Status
Move Media
Setup
Functions
History

I think that is not the nsr web, isn't?

Thank you very much

[Torsten.]

The NSR (network storage router) is the card next to the library controller. The default is DHCP - name "Router".

You can also connect the serial cable to a PC's COM port and connect via terminal emulator (e.g. putty) with 115200 as speed.

[i5513]

Hi,

I tried to sniff with a wire dhcp petitions (turning off and turning on NSR), but I couldn't get any packet.

Too I tried to go to 1.1.1.1 (default ip but it didn't reply)

So I guess my only posibility to update the firmware is using a serial wire, which I haven't in this moment. I'll wait to have one to continue...

If is there any possibility to continue without serial wire, please tell me which .

Thank you very much.

[Torsten.]

As said, with newer firmware (since some years) the default setting is DHCP (name: Router). If the deice did not get an IP, you cannot access via network.

The serial cable was delivered with the device, have a look.

[i5513]

I connected through hyperterminal / minicom and no response from nsr.

I configured the connexion like says the manual.

Any tip more ?

Thanks

[Torsten.]

Did you use the correct speed setting and the _correct_ cable? The cable is different from the MSL cable!

[i5513]

I connected with a usb-serial cable to other cable "connector which I don't known how it is named in msr"-serial.

One extreme was male and the other female.

The speed was 115200

Thank you

[Torsten.]

3 possibilities:

1) still wrong canle
2) wrong speed (maybe modified)
3) NSR is frozen

[i5513]

Maybe I'm using a not valid manual.

Do you have the link to it ?

Thanks

[i5513]

Which mode is ¿full or null?

Thanks

[Torsten.]

full or null?

[i5513]

Sorry, my mistake.

I wanted to say "cross" or "direct" cable.

Thanks

[Torsten.]

There are 2 cables delivered with the lib - both look very similar. They are RJ11 on both sides + a small converter box.

One is for the lib controller, the other is for the NSR. The pinout is different.

[i5513]

I understand then, I need a "direct" usb-serial cable, isn't it?

[Torsten.]

The "device offline" message may point to a problem with the NSR.

However, if no network or serial connection possible now, I would try to discover the device with LTT and update the firmware.

BTW, the web interface and the FTP interface will give access to the config in order to save - you should consider to do this.

[i5513]

Hi,

Do you mean http://h18000.www1.hp.com/products/storageworks/ltt/index.html ?

[Torsten.]

Exactly.

If your NSR is presented to outside, LTT is able to update.

[i5513]

We are going to call hp support, so they can came here and tell us how to fix the problem.

I'll post the final solution

Thank you very much