For more details, please see ourCookie Policy.


Fibre Channel (SAN)

Reply
Valued Contributor
Posts: 931
Registered: ‎12-30-2009

Re: Need help, ESX paths all showing dead

yes that would make it work.

if you want to be consisted you'll first create an alias for that wwn and add the alias to the zoning.

update: picture shows you added the wwN(ode)n. If the storage is still not visible add the wwP(ort)n.

its the one below the one you've added (on the image)

Valued Contributor
Posts: 931
Registered: ‎12-30-2009

Re: Need help, ESX paths all showing dead

Ask yourself the question how to lose data if I don't have any LUN's. Some data has already been lost, it happend when all the paths went dead.

You'll lose more data if you zone improperly and IF a host (or operator) should initialize/format the LUN.

ESX will not do this. Just rescan (check the box search for VMFS or something)

Contributor
Posts: 22
Registered: ‎03-30-2011

Re: Need help, ESX paths all showing dead

Thank you Dion!  I added the port to the zone and enabled it.  LUNS are back and everything appears to be working again.

Can I mirror the config on the second brocade using the second ESX storage adaptors?

Valued Contributor
Posts: 931
Registered: ‎12-30-2009

Re: Need help, ESX paths all showing dead

Yes you should do the same for your second fabric.

When done you'll have a second path for each LUN, please observe loadbalancing options.

This should by MRU for Active Passive array's and Fixed for Active Active array's.

I think Clariion is active active but i'm not sure, so check with your vendor.

Contributor
Posts: 22
Registered: ‎03-30-2011

Re: Need help, ESX paths all showing dead

Cool.

I apprecaite your help.

Valued Contributor
Posts: 931
Registered: ‎12-30-2009

Re: Need help, ESX paths all showing dead

NP glad you were able to fix it, now go book your courses

Contributor
Posts: 22
Registered: ‎03-30-2011

Re: Need help, ESX paths all showing dead

One other question, the one host still can't see the luns.  Navisphere doesnt see the host either so something is misconfigured, just not sure what.  Not sure if you can help me with this though.  It seems as though its connected to the onboard storage instead of the SAN?  Its not clear to me.

Alias 'hp_blade6_port2'  (part of zone config as well)

Device port WWN: xxxx:16

clariion4.PNG

ESX host vmba0: xxx:14

esx.PNG

ESX host vmhba1: xxxx:16

esx1.PNG

Valued Contributor
Posts: 931
Registered: ‎12-30-2009

Re: Need help, ESX paths all showing dead

One other question, the one host still can't see the luns.  Navisphere  doesnt see the host either so something is misconfigured, just not sure  what.  Not sure if you can help me with this though.  It seems as though  its connected to the onboard storage instead of the SAN?  Its not clear  to me.

Take a closer look on the images you've attached. To be precise the last 2 images of your reply.

In both images you've selected the fibre channel HBA and you see a LUN hanging of vmhba1 instead of vmhba2.

vmhba2 is your onboard (local to the blade only) e200 raid controller.

If that doesn't convice you, take a look at identifier on the the last image.

after naa. you see a lot of number identifing a LUN. If you look more closely you'll see its from the Clariion because the first 16 numbers are the Clariion's WWN.

So blade6 sees a LUN from the Clariion, but only on one path and perhaps the wrong one.

In the fabric that belongs to vmhba0 you can do an portshow (CLI) command againt the storage port and hostport.

Do you see the expected wwn's below devices connected and is/has the port status online/insync?

With the wwn's from portshow, can you do an fcping wwn1 wwn2 to figure out if you get an response and if it thinks its zoned (3rd line from the top after command being given, but I'm guessing a bit)

Contributor
Posts: 22
Registered: ‎03-30-2011

Re: Need help, ESX paths all showing dead

Thanks again, that makes sense.

One other question.

I think you mentioned this before.

All of my initiators are in one zone, that is not best practice correct?  They should all be in separate zones? I would then end up with multiple zones each containting one single alias  and all of which are part of one zone config?  So instead of one zone in my zone config file I would have 7?

ZONE CONFIG:

Zone:

- zone_1

     -alias1

-zone_2

     -alias2

-zone_3

     -alias3

...


VS:

ZONE CONFIG:

Zone:

-zone_1

     -alias1

     -alias2

     -alias3

.....

Valued Contributor
Posts: 931
Registered: ‎12-30-2009

Re: Need help, ESX paths all showing dead

You forgot to add the storage array.

ZONE CONFIG:

Zone:

- zone_1

     -blade_alias1

     -Clariion_Alias

-zone_2

     -blade_alias2

     -Clariion_Alias

-zone_3

     -blade_alias3

     -Clariion_Alias

Etc etc.

HP design reference guide can be found here http://bizsupport2.austin.hp.com/bc/docs/support/SupportManual/c00403562/c00403562.pdf

Perhaps EMC has a best practice for your Clariion but you'll have to log on at powerlink.emc.com

Join the Broadcom Community

Get quick and easy access to valuable resources across the Broadcom Community Network.