TruCluster
cancel
Showing results for 
Search instead for 
Did you mean: 

Help me with following message.

Tim_SW
Advisor

Help me with following message.

Why I can to shruggle with

mchan0: Resetting controller, come_online = 0
mchan0: mchan_reset has taken controller offline
panic (cpu 0): rm_release_global_tag_hold

?
Why does it appears? Offer anybody the ways of resolution, please.

Thanks to all, who can help.
6 REPLIES
Martin Moore
HPE Pro

Re: Help me with following message.

That looks like a memory channel hardware problem. Tell us more about your cluster; is it a 2-member cluster running in virtual hub mode, or something else? What Alpha models are the systems involved? What version of Tru64 UNIX are they running?

Martin
I work for HP
A quick resolution to technical issues for your HP Enterprise products is just a click away HP Support Center Knowledge-base
See Self Help Post for more details

Tim_SW
Advisor

Re: Help me with following message.

Yes, I use CCMAB II, Alpha ES40, TruCluster 1.6, DigitalUnix 4.0F, connection is in virtual hub mode (2 computers).
Martin Moore
HPE Pro

Re: Help me with following message.

OK, in that case you could try running the memory channel diagnostics. To do this, bring both systems down and at the console prompt, run the 'mc_diag' and 'mc_cable' commands. Note: 'mc_cable' needs to run at the same time on both systems to produce meaningful results.
I work for HP
A quick resolution to technical issues for your HP Enterprise products is just a click away HP Support Center Knowledge-base
See Self Help Post for more details

Tim_SW
Advisor

Re: Help me with following message.

Martin, how it isn't strangely, my SRM console hasn't this service. It's in considerable part make big difficults! I don't know what to do else.
Martin Moore
HPE Pro

Re: Help me with following message.

That is indeed strange. Those commands are part of the SRM (firmware) command set.

Are you trying this with the systems shut all the way down, i.e., halted and at the P00>>> prompt? That's the only time the commands are accessible.

I work for HP
A quick resolution to technical issues for your HP Enterprise products is just a click away HP Support Center Knowledge-base
See Self Help Post for more details

cnb
Honored Contributor

Re: Help me with following message.

Hi Tim,

From your other posts, it appears that you are running in VH mode, but the controllers are jumpered for Standard Hub mode?

mchan0: jumpered as HUB configuration
mchan0: Probe successful
mchan0 at pci1 slot 2
mchan0: controller_state = 10 node_state = 0 ******node_id = ffffffff****

Check your jumper settings on both adapters, if going from node to node, then set them for VH mode and one adapter is addressed as node 0 and the other adapter is addressed as node 1. If you are using a pysical CCMHB Hub interconnect then set both adapters for a physical hub connection:

J1 - Hub Mode
J1 is used to configure the module for one of three modes of operation depending on
the cluster configuration. In virtual hub (VH) mode, two systems are cabled together
directly without a hub. In VH mode, the J1 jumper on one CCMAB adapter must be set to Virtual Hub Node 0 (VH0) and the J1 jumper on the other CCMAB to Virtual Hub Node 1 (VH1). The J1 setting determines the node ID.

If the module is connected to a CCMHB hub (standard hub mode), then it is configured in standard (STD) mode and all CCMAB adapters must have the same J1 jumper configuration. If you are upgrading from a two-node virtual hub to a standard
mode configuration with a CCMHB, check and change the J1 jumpers on all CCMAB adapters.

Node 1: No jumpers set (VH1)
Node 0: Jumper middle and right pin (VH0)
Virtual Hub Mode (no hub present)
All Nodes: Jumper middle and left pins
Standard Hub Mode (with hub hardware) 1 2 3 connection

hth,