HPE Storage Users Group

Hello.

I can't finish the Node Rescue process.

It is a 3PAR 72000 MU6.

I am installing a new node which is OK. It performs the entire process but fails due to "space" issues. I have made several node changes and this is the first time it has happened to me.

The IDE disk is the same as the other NODE (blue).

I have run all known processes from the WACK>



In short, I follow all the steps and execute:

whack>unset sys_serial

whack>ahci init

Whack>ide erase

whack>reset



Everything runs ok, but never joins the cluster and ends with this error

[ 45.018834] Total CM for this node is 4 GB
[ 45.018838] patrol_delay 4614 msecs
[ 45.018850] Starting eagle_cm_patrol_thread....
[ 45.018853] Num Srcs 14 Src Size 16384
[ 45.060562] Starting scan mir log thread 0x226b
[ 45.060564] Setting recovery log scan delay to 292 ms
init: Loading tpd kernel module: done.
init: Remapping interrupts
init: WARNING: irq name contains comma delimited list [aerdrv 0x70, aerdrv 0x70, aerdrv 0x70, aerdrv 0x70, aerdrv 0x70, aerdrv 0x70, aerdrv]
init: WARNING: irq name contains comma delimited list [nemoe 0x78, Eos intr]
init: Loading tpd kernel patch modules: Patch99 Patch110 Patch114 P119 Patch135 Patch138 Patch139 Patch142 Patch149 Patch160 Patch165
init: Configuring powerfail device /dev/pfail_storage: 801
init: done.
init: Initializing eanet0: done.
init: Rotating PM logs: done.
init: Rotating PR event logs: done.
init: Starting Process Manager: pm.
init: Rotating pr_sync_ide logs: done.
init: Starting PR Sync IDE: pr_sync_ide.
init: coreman: tpdinit run at epoch=1664819228sec with boot at epoch=1664819176sec
init: coreman: file system usage - /dev/sda7 - used=15227584KB (used-core=172KB) free=46257040KB total=61484624KB
init: coreman: space removal targets - freeMin=0KB freePct=0KB corePct=0KB targetToRemove=0KB targetToRemoveMinLow=0KB
init: coreman: exit - no free space requirements to satisfy
init: Launching tpd_cfg_tornado_dcncage script to work around 'Configures a Tornado Enclosures DP-1 port' issue
init: Running l
3PAR(R) InForm(tm) OS 3.2.2.709 1612423-1 ttyS0
1612XXX-1 login:

I Put the log file here.
I do not have current support from HPE.
I will greatly appreciate the help.



Thanks!!!!
Sebastian.

I see one important line in there…

[ 42.647579] Node type mismatch. Node 0 should be of type 2 node value, but the node type that was received was 2 node enhanced.Node 0 cannot join the cluster
[ 42.675598] Remote node type mismatch. Node 0 should be of type 2 node value, but the node type that was received was 2 node enhanced.Node 0 cannot join the cluster
[

From the looks of it you have a 7200c replacement node in a 7200 system…. or the other way around.

MammaGutt wrote:I see one important line in there…

[ 42.647579] Node type mismatch. Node 0 should be of type 2 node value, but the node type that was received was 2 node enhanced.Node 0 cannot join the cluster
[ 42.675598] Remote node type mismatch. Node 0 should be of type 2 node value, but the node type that was received was 2 node enhanced.Node 0 cannot join the cluster
[

From the looks of it you have a 7200c replacement node in a 7200 system…. or the other way around.

Thanks for your answer.
Are the nodes not compatible with each other?
This compatibility is not achieved with the procedure in the Wack> ?

From Whack> on the replacement node issue prom edit <CR>. Then hit return to see each item in the prom.
SMI Whack>prom edit
Board Spin: 03
Size * 256 bytes: 04
Board Class: 920
Board Base: 200040
Board Rev: 91
Assembly Vendor: FXN
Assembly Year: 2012
Assembly Week: 30
Assembly Day: 03
Assembly Serial: 00103530
Alternate Class: ffff
Alternate Base: ffff
Alternate Dash: ff
System Serial: 9902109
Node ID: 03 FF or value of slot node is installed in
Midplane Type: 1b See Below for current settings.
Node Type: 40
SMI Whack>
Midplane and Node Type must match the other nodes. This list will change as new models are released . Current selections as of August 2014 are
StoreServ 7200: Mid-plane Type 1a Node Type 20
StoreServ 7400: Mid-plane Type 1b Node Type 40
StoreServ 7450: Mid-plane Type 1e Node Type 80

sebastian7780_2 wrote:

MammaGutt wrote:I see one important line in there…

[ 42.647579] Node type mismatch. Node 0 should be of type 2 node value, but the node type that was received was 2 node enhanced.Node 0 cannot join the cluster
[ 42.675598] Remote node type mismatch. Node 0 should be of type 2 node value, but the node type that was received was 2 node enhanced.Node 0 cannot join the cluster
[

From the looks of it you have a 7200c replacement node in a 7200 system…. or the other way around.

Thanks for your answer.
Are the nodes not compatible with each other?
This compatibility is not achieved with the procedure in the Wack> ?

From Whack> on the replacement node issue prom edit <CR>. Then hit return to see each item in the prom.
SMI Whack>prom edit
Board Spin: 03
Size * 256 bytes: 04
Board Class: 920
Board Base: 200040
Board Rev: 91
Assembly Vendor: FXN
Assembly Year: 2012
Assembly Week: 30
Assembly Day: 03
Assembly Serial: 00103530
Alternate Class: ffff
Alternate Base: ffff
Alternate Dash: ff
System Serial: 9902109
Node ID: 03 FF or value of slot node is installed in
Midplane Type: 1b See Below for current settings.
Node Type: 40
SMI Whack>
Midplane and Node Type must match the other nodes. This list will change as new models are released . Current selections as of August 2014 are
StoreServ 7200: Mid-plane Type 1a Node Type 20
StoreServ 7400: Mid-plane Type 1b Node Type 40
StoreServ 7450: Mid-plane Type 1e Node Type 80

Nodes need to be indentical. 7200 and 7200c are not (7200c has more CPU cores and I believe also control/data cache)

Mammagutt

You were right,.
The problem has already been resolved.
7200 node was being used in a 3PAR 7200c.
all ok!
Thank you very much.