Discussion:
[xcat-user] xcat.genesis.doxcat: Unrecognized directive (dest=)
Nicolas Roosen
2017-07-04 16:36:22 UTC
Permalink
Hello,

last week I was able to discover sequentially 4 nodes.

Today I try to discover a new one, using the same set of commands.

But then I get this on the console:

Jul 4 18:17:10 10.33.241.3 (none) rsyslogd: [origin
software="rsyslogd" swVersion="5.8.10" x-pid="2361"
x-info="http://www.rsyslog.com"] start
Jul 4 18:17:10 10.33.241.3 (none) xcat.genesis.doxcat: Getting initial
certificate --> 10.33.240.1:3001
Jul 4 18:17:30 10.33.241.3 (none) xcat.genesis.doxcat: Running
getdestiny --> 10.33.240.1:3001
Jul 4 18:17:50 10.33.241.3 (none) xcat.genesis.doxcat: Received destiny=
Jul 4 18:17:50 10.33.241.3 (none) xcat.genesis.doxcat: The destiny=,
destiny parameters=
Jul 4 18:17:50 10.33.241.3 (none) xcat.genesis.doxcat: Unrecognized
directive (dest=)
Jul 4 18:17:54 10.33.241.3 (none) xcat.genesis.doxcat: ... Will retry
xCAT in 90 seconds
Jul 4 18:18:04 10.33.241.3 (none) xcat.genesis.doxcat: ... Will retry
xCAT in 80 seconds
Jul 4 18:18:14 10.33.241.3 (none) xcat.genesis.doxcat: ... Will retry
xCAT in 70 seconds
Jul 4 18:18:24 10.33.241.3 (none) xcat.genesis.doxcat: ... Will retry
xCAT in 60 seconds

I tried different settings, but cannot find out why the node doesn't get
a discover destiny parameter?


Here is what I've done:

* mkdef -t node -o n043 groups="all,ipmi,node-giga" arch=x86_64 mgt=ipmi
bmc=10.33.192.57 mac=0c:c4:7a:71:f1:17 ip=10.33.240.57
chain="runcmd=bmcsetup,osimage=centos7.3-x86_64-node-giga"
provmethod=centos7.3-node-giga

* nodediscoverstart noderange=n043

Then booted the node via IPMI ...


Any thoughts?
Thanks.

--
Nicolas
Xiao Peng Wang
2017-07-05 01:23:19 UTC
Permalink
Please show out the logs in MN during the discovery happens.

Using IBM Verse, send from my iPhone.

圚 2017幎7月5日䞊午12:37:37***@hpe.com 写道

From: ***@hpe.com
To: xcat-***@lists.sourceforge.net
Cc:
Date: 2017幎7月5日 䞊午12:37:37
Subject: [xcat-user] xcat.genesis.doxcat: Unrecognized directive (dest=)


Hello,
last week I was able to discover sequentially 4 nodes.
Today I try to discover a new one, using the same set of commands.
But then I get this on the console:
Jul 4 18:17:10 10.33.241.3 (none) rsyslogd: [origin
software="rsyslogd" swVersion="5.8.10" x-pid="2361"
x-info="http://www.rsyslog.com"] start
Jul 4 18:17:10 10.33.241.3 (none) xcat.genesis.doxcat: Getting initial
certificate --> 10.33.240.1:3001
Jul 4 18:17:30 10.33.241.3 (none) xcat.genesis.doxcat: Running
getdestiny --> 10.33.240.1:3001
Jul 4 18:17:50 10.33.241.3 (none) xcat.genesis.doxcat: Received destiny=
Jul 4 18:17:50 10.33.241.3 (none) xcat.genesis.doxcat: The destiny=,
destiny parameters=
Jul 4 18:17:50 10.33.241.3 (none) xcat.genesis.doxcat: Unrecognized
directive (dest=)
Jul 4 18:17:54 10.33.241.3 (none) xcat.genesis.doxcat: ... Will retry
xCAT in 90 seconds
Jul 4 18:18:04 10.33.241.3 (none) xcat.genesis.doxcat: ... Will retry
xCAT in 80 seconds
Jul 4 18:18:14 10.33.241.3 (none) xcat.genesis.doxcat: ... Will retry
xCAT in 70 seconds
Jul 4 18:18:24 10.33.241.3 (none) xcat.genesis.doxcat: ... Will retry
xCAT in 60 seconds
I tried different settings, but cannot find out why the node doesn't get
a discover destiny parameter?
Here is what I've done:
* mkdef -t node -o n043 groups="all,ipmi,node-giga" arch=x86_64 mgt=ipmi
bmc=10.33.192.57 mac=0c:c4:7a:71:f1:17 ip=10.33.240.57
chain="runcmd=bmcsetup,osimage=centos7.3-x86_64-node-giga"
provmethod=centos7.3-node-giga
* nodediscoverstart noderange=n043
Then booted the node via IPMI ...
Any thoughts?
Thanks.
--
Nicolas
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
xCAT-user mailing list
xCAT-***@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/xcat-user
Nicolas Roosen
2017-07-05 07:49:46 UTC
Permalink
Hi,

here are some logs.

So first we removed an unwanted server from the discovery process:

Jul 4 10:42:02 adm xcat[2941]: xcatd: Processing discovery request from
10.33.241.181
Jul 4 10:42:02 adm xcat[2941]: xcat.discovery.aaadiscovery:
(0c:c4:7a:70:e0:80) Got a discovery request, attempting to discover the
node...
Jul 4 10:42:02 adm xcat[2941]: xcat.discovery.blade:
(0c:c4:7a:70:e0:80) Warning: Could not find any nodes using blade-based
discovery
Jul 4 10:42:02 adm xcat[2941]: xcat.discovery.switch:
(0c:c4:7a:70:e0:80) Warning: Could not find any nodes using switch-based
discovery
Jul 4 10:42:02 adm xcat[2941]: xcat.discovery.mtms: (0c:c4:7a:70:e0:80)
Warning: Could not find any node for Supermicro:*A00AD6F3 using
mtms-based discovery
Jul 4 10:42:02 adm xcat[2941]: xcat.discovery.zzzdiscovery:
(0c:c4:7a:70:e0:80) Failed to discover the node.

Jul 4 10:43:49 adm xcat[2820]: xCAT: Allowing nodediscoverls for root
from localhost
Jul 4 10:44:25 adm xcat[2843]: xCAT: Allowing nodediscoverdef -h for
root from localhost
Jul 4 10:44:32 adm xcat[2852]: xCAT: Allowing nodediscoverdef -r -u
00000000-0000-0000-0000-0CC47A70E080 for root from localhost
Jul 4 10:44:35 adm xcat[2859]: xCAT: Allowing nodediscoverls for root
from localhost
Jul 4 10:44:39 adm xcat[2869]: xCAT: Allowing nodediscoverstatus for
root from localhost
Jul 4 10:45:25 adm xcat[2889]: xCAT: Allowing nodels for root from
localhost

The we tried to create a new one, but couldn't see any discovery process
information in the logs:

Jul 4 16:02:23 adm xcat[6639]: DEBUG xcatd: open new process : xcatd
SSL: mkdef for ***@localhost
Jul 4 16:02:23 adm xcat[6639]: xCAT: Allowing mkdef -t node -o n043
groups=all,ipmi,node-giga arch=x86_64 mgt=ipmi bmc=172.30.192.57
mac=0c:c4:7a:71:f1:17 ip=172.30.240.57
chain=runcmd=bmcsetup,osimage=centos7.3-x86_64-node-giga
provmethod=centos7.3-node-giga for root from localhost
Jul 4 16:02:23 adm xcat[6640]: DEBUG xcatd: dispatch request 'mkdef -t
node -o n043 groups=all,ipmi,node-giga arch=x86_64 mgt=ipmi
bmc=172.30.192.57 mac=0c:c4:7a:71:f1:17 ip=172.30.240.57
chain=runcmd=bmcsetup,osimage=centos7.3-x86_64-node-giga
provmethod=centos7.3-node-giga' to plugin 'DBobjectdefs'
Jul 4 16:02:23 adm xcat[6640]: DEBUG xcatd: handle request 'mkdef' by
plugin 'DBobjectdefs''s process_request
Jul 4 16:02:23 adm xcat[6639]: DEBUG xcatd: close connection with
***@localhost
Jul 4 16:02:29 adm xcat[6660]: DEBUG xcatd: connection from ***@localhost
Jul 4 16:02:29 adm xcat[6660]: DEBUG xcatd: open new process : xcatd
SSL: lsdef for ***@localhost
Jul 4 16:02:29 adm xcat[6660]: xCAT: Allowing lsdef n043 for root from
localhost
Jul 4 16:02:29 adm xcat[6661]: DEBUG xcatd: dispatch request 'lsdef
n043' to plugin 'DBobjectdefs'
Jul 4 16:02:29 adm xcat[6661]: DEBUG xcatd: handle request 'lsdef' by
plugin 'DBobjectdefs''s process_request
Jul 4 16:02:29 adm xcat[6660]: DEBUG xcatd: close connection with
***@localhost
Jul 4 16:02:58 adm xcat[6680]: DEBUG xcatd: connection from ***@localhost
Jul 4 16:02:58 adm xcat[6680]: DEBUG xcatd: open new process : xcatd
SSL: nodediscoverstart for ***@localhost
Jul 4 16:02:58 adm xcat[6681]: DEBUG xcatd: dispatch request
'nodediscoverstart noderange=n043' to plugin 'seqdiscovery'
Jul 4 16:02:58 adm xcat[6681]: DEBUG xcatd: handle request
'nodediscoverstart' by plugin 'seqdiscovery''s process_request
Jul 4 16:02:58 adm xcat[6680]: xCAT: Allowing nodediscoverstart
noderange=n043 for root from localhost
Jul 4 16:02:58 adm xcat[6682]: DEBUG xcatd: dispatch request
'nodediscoverstart noderange=n043' to plugin 'profilednodes'
Jul 4 16:02:58 adm xcat[6682]: DEBUG xcatd: handle request
'nodediscoverstart' by plugin 'profilednodes''s process_request
Jul 4 16:02:58 adm xcat[6680]: DEBUG xcatd: close connection with
***@localhost
Jul 4 16:03:01 adm xcat[6688]: DEBUG xcatd: connection from ***@localhost
Jul 4 16:03:01 adm xcat[6688]: DEBUG xcatd: open new process : xcatd
SSL: lsdef for ***@localhost
Jul 4 16:03:01 adm xcat[6688]: xCAT: Allowing lsdef n043 for root from
localhost
Jul 4 16:03:01 adm xcat[6689]: DEBUG xcatd: dispatch request 'lsdef
n043' to plugin 'DBobjectdefs'
Jul 4 16:03:01 adm xcat[6689]: DEBUG xcatd: handle request 'lsdef' by
plugin 'DBobjectdefs''s process_request
Jul 4 16:03:01 adm xcat[6688]: DEBUG xcatd: close connection with
***@localhost


I might miss something, but can't find out what.

Is the "bmcdiscover" absolutely mandatory in a sequential discovery process?


Thanks.
Nicolas

On 07/05/2017 03:23 AM, Xiao Peng Wang wrote:
> Please show out the logs in MN during the discovery happens.
>
> Using IBM Verse, send from my iPhone.
>
> ------------------------------------------------------------------------
> 在 2017年7月5日,上午12:37:37,***@hpe.com 写道:
>
> From: ***@hpe.com
> To: xcat-***@lists.sourceforge.net
> Cc:
> Date: 2017年7月5日 上午12:37:37
> Subject: [xcat-user] xcat.genesis.doxcat: Unrecognized directive (dest=)
>
> Hello,

[...]

--
Nicolas
Xiao Peng Wang
2017-07-05 14:42:43 UTC
Permalink
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
Nicolas Roosen
2017-07-05 15:32:22 UTC
Permalink
Hi,

On 07/05/2017 04:42 PM, Xiao Peng Wang wrote:
> No, bmcdiscovery is not mandatory for sequential discovery.
> The correct procedure should be:
> 1. start sequential discovery first: nodediscoverstart
> 2. boot the node, then you could see the following messages
> xcat.discovery.seqdiscovery
> But in your case that there was no any discovery messages after starting
> 'nodediscoverstart', so you need check the status of the your nodes.
>

Yes, so I tried the nodediscoverstart with an IP range instead of just
the single node hostname:

# nodediscoverstart noderange=172.30.241.2-200 -V
Sequential Discovery: Started:
Number of free node names: 1

====================Free Nodes===================
NODE HOST IP BMC IP
172.30.241.2-200 --no free-- --no free--


If I understand correctly, there are no free nodes available for
discover, which might be my problem?

On the node side console, the only logs I have are the bellow:

Jul 5 17:22:51 172.30.241.69 (none) xcat.genesis.doxcat: Getting
initial certificate --> 172.30.240.1:3001
Jul 5 17:23:11 172.30.241.69 (none) xcat.genesis.doxcat: Running
getdestiny --> 172.30.240.1:3001
Jul 5 17:23:31 172.30.241.69 (none) xcat.genesis.doxcat: Received destiny=
Jul 5 17:23:31 172.30.241.69 (none) xcat.genesis.doxcat: The destiny=,
destiny parameters=
Jul 5 17:23:31 172.30.241.69 (none) xcat.genesis.doxcat: Unrecognized
directive (dest=)
Jul 5 17:23:37 172.30.241.69 (none) xcat.genesis.doxcat: ... Will
retry xCAT in 210 seconds
Jul 5 17:23:47 172.30.241.69 (none) xcat.genesis.doxcat: ... Will
retry xCAT in 200 seconds
Jul 5 17:23:57 172.30.241.69 (none) xcat.genesis.doxcat: ... Will
retry xCAT in 190 seconds
Jul 5 17:24:07 172.30.241.69 (none) xcat.genesis.doxcat: ... Will
retry xCAT in 180 seconds



> Best Regards
> ----------------------------------------------------------------------
> Wang Xiaopeng (王晓朋)
> Manager for HPC SW Dev: xCAT, ESSL, SMI, Test
> IBM China Systems Laboratory (CSL)
>
> Tel: 86-10-82453455
> Email: ***@cn.ibm.com
>
> ----- Original message -----
> From: Nicolas Roosen <***@hpe.com>
> To: <xcat-***@lists.sourceforge.net>
> Cc:
> Subject: Re: [xcat-user] 回复: xcat.genesis.doxcat: Unrecognized
> directive (dest=)
> Date: Wed, Jul 5, 2017 3:53 PM
> Hi,
>
> here are some logs.
>
> So first we removed an unwanted server from the discovery process:
>

[...]

>
>
> I might miss something, but can't find out what.
>
> Is the "bmcdiscover" absolutely mandatory in a sequential discovery
> process?
>
>
> Thanks.
> Nicolas
>
> On 07/05/2017 03:23 AM, Xiao Peng Wang wrote:
> > Please show out the logs in MN during the discovery happens.
> >
> > Using IBM Verse, send from my iPhone.
> >
> >
> ------------------------------------------------------------------------
> > 在 2017年7月5日,上午12:37:37,***@hpe.com 写道:
> >
> > From: ***@hpe.com
> > To: xcat-***@lists.sourceforge.net
> > Cc:
> > Date: 2017年7月5日 上午12:37:37
> > Subject: [xcat-user] xcat.genesis.doxcat: Unrecognized directive
> (dest=)
> >
> > Hello,
>
> [...]
>
> --
> Nicolas
>

--
Nicolas
Xiao Peng Wang
2017-07-06 01:29:42 UTC
Permalink
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
Nicolas Roosen
2017-07-06 08:30:15 UTC
Permalink
On 07/06/2017 03:29 AM, Xiao Peng Wang wrote:
> 1. please be sure that you can see the message like
> 'xcat.discovery.seqdiscovery' in your MN log. This is important to know
> that the sequential discovery process has been triggered.

That's my problem I think, I tried this morning again, and although the
"nodediscoverstart" seems to be running, there are *no* messages into
the logs (debug is set to 2).

I'd like to debug this further but I have no clue where to look at.


nodediscoverstart -V noderange=n042
Sequential Discovery: Started:
Number of free node names: 0

====================Free Nodes===================
NODE HOST IP BMC IP


* /var/log/messages


Jul 6 10:15:41 adm xcat[16015]: DEBUG xcatd: connection from ***@localhost
Jul 6 10:15:41 adm xcat[16015]: DEBUG xcatd: open new process : xcatd
SSL: nodediscoverstart for ***@localhost
Jul 6 10:15:41 adm xcat[16016]: DEBUG xcatd: dispatch request
'nodediscoverstart -V noderange=n042' to plugin 'seqdiscovery'
Jul 6 10:15:41 adm xcat[16016]: DEBUG xcatd: handle request
'nodediscoverstart' by plugin 'seqdiscovery''s process_request
Jul 6 10:15:41 adm xcat[16015]: xCAT: Allowing nodediscoverstart -V
noderange=n042 for root from localhost
Jul 6 10:15:41 adm xcat[16017]: DEBUG xcatd: dispatch request
'nodediscoverstart -V noderange=n042' to plugin 'profilednodes'
Jul 6 10:15:41 adm xcat[16017]: DEBUG xcatd: handle request
'nodediscoverstart' by plugin 'profilednodes''s process_request
Jul 6 10:15:41 adm xcat[16015]: DEBUG xcatd: close connection with
***@localhost


* /var/log/xcat/cluster.log


Jul 6 10:15:41 adm xcat[16015]: DEBUG xcatd: connection from ***@localhost
Jul 6 10:15:41 adm xcat[16015]: DEBUG xcatd: open new process : xcatd
SSL: nodediscoverstart for ***@localhost
Jul 6 10:15:41 adm xcat[16016]: DEBUG xcatd: dispatch request
'nodediscoverstart -V noderange=n042' to plugin 'seqdiscovery'
Jul 6 10:15:41 adm xcat[16016]: DEBUG xcatd: handle request
'nodediscoverstart' by plugin 'seqdiscovery''s process_request
Jul 6 10:15:41 adm xcat[16015]: xCAT: Allowing nodediscoverstart -V
noderange=n042 for root from localhost
Jul 6 10:15:41 adm xcat[16017]: DEBUG xcatd: dispatch request
'nodediscoverstart -V noderange=n042' to plugin 'profilednodes'
Jul 6 10:15:41 adm xcat[16017]: DEBUG xcatd: handle request
'nodediscoverstart' by plugin 'profilednodes''s process_request
Jul 6 10:15:41 adm xcat[16015]: DEBUG xcatd: close connection with
***@localhost

* /var/log/xcat/subcallingtrace is attached in any case.


Thanks.
Nicolas

> 2. The command 'nodediscoverstart noderange=' only supports a real
> noderange instea of IP ranges. so you need to make this command work as
> a charm before the discovery start.
>
> Best Regards
> ----------------------------------------------------------------------
> Wang Xiaopeng (王晓朋)
> Manager for HPC SW Dev: xCAT, ESSL, SMI, Test
> IBM China Systems Laboratory (CSL)
>
> Tel: 86-10-82453455
> Email: ***@cn.ibm.com
>
> ----- Original message -----
> From: Nicolas Roosen <***@hpe.com>
> To: <xcat-***@lists.sourceforge.net>
> Cc:
> Subject: Re: [xcat-user] 回倍 xcat.genesis.doxcat: Unrecognized
> directive (dest=)
> Date: Wed, Jul 5, 2017 11:34 PM
> Hi,
>
> On 07/05/2017 04:42 PM, Xiao Peng Wang wrote:
> > No, bmcdiscovery is not mandatory for sequential discovery.
> > The correct procedure should be:
> > 1. start sequential discovery first: nodediscoverstart
> > 2. boot the node, then you could see the following messages
> > xcat.discovery.seqdiscovery
> > But in your case that there was no any discovery messages after
> starting
> > 'nodediscoverstart', so you need check the status of the your nodes.
> >
>
> Yes, so I tried the nodediscoverstart with an IP range instead of just
> the single node hostname:
>
> # nodediscoverstart noderange=172.30.241.2-200 -V
> Sequential Discovery: Started:
> Number of free node names: 1
>
> ====================Free Nodes===================
> NODE HOST IP BMC IP
> 172.30.241.2-200 --no free-- --no free--
>
>
> If I understand correctly, there are no free nodes available for
> discover, which might be my problem?
>
> On the node side console, the only logs I have are the bellow:
>
> Jul 5 17:22:51 172.30.241.69 (none) xcat.genesis.doxcat: Getting
> initial certificate --> 172.30.240.1:3001
> Jul 5 17:23:11 172.30.241.69 (none) xcat.genesis.doxcat: Running
> getdestiny --> 172.30.240.1:3001
> Jul 5 17:23:31 172.30.241.69 (none) xcat.genesis.doxcat: Received
> destiny=
> Jul 5 17:23:31 172.30.241.69 (none) xcat.genesis.doxcat: The destiny=,
> destiny parameters=
> Jul 5 17:23:31 172.30.241.69 (none) xcat.genesis.doxcat: Unrecognized
> directive (dest=)
> Jul 5 17:23:37 172.30.241.69 (none) xcat.genesis.doxcat: ... Will
> retry xCAT in 210 seconds
> Jul 5 17:23:47 172.30.241.69 (none) xcat.genesis.doxcat: ... Will
> retry xCAT in 200 seconds
> Jul 5 17:23:57 172.30.241.69 (none) xcat.genesis.doxcat: ... Will
> retry xCAT in 190 seconds
> Jul 5 17:24:07 172.30.241.69 (none) xcat.genesis.doxcat: ... Will
> retry xCAT in 180 seconds
>
>
>
> > Best Regards
> >
> ----------------------------------------------------------------------
> > Wang Xiaopeng (王晓朋)
> > Manager for HPC SW Dev: xCAT, ESSL, SMI, Test
> > IBM China Systems Laboratory (CSL)
> >
> > Tel: 86-10-82453455
> > Email: ***@cn.ibm.com
> >
> > ----- Original message -----
> > From: Nicolas Roosen <***@hpe.com>
> > To: <xcat-***@lists.sourceforge.net>
> > Cc:
> > Subject: Re: [xcat-user] 回倍 xcat.genesis.doxcat: Unrecognized
> > directive (dest=)
> > Date: Wed, Jul 5, 2017 3:53 PM
> > Hi,
> >
> > here are some logs.
> >
> > So first we removed an unwanted server from the discovery
> process:
> >
>
> [...]
>
> >
> >
> > I might miss something, but can't find out what.
> >
> > Is the "bmcdiscover" absolutely mandatory in a sequential
> discovery
> > process?
> >
> >
> > Thanks.
> > Nicolas
> >
> > On 07/05/2017 03:23 AM, Xiao Peng Wang wrote:
> > > Please show out the logs in MN during the discovery happens.
> > >
> > > Using IBM Verse, send from my iPhone.
> > >
> > >
> >
> ------------------------------------------------------------------------
> > > 圚 2017幎7月5日䞊午12:37:37***@hpe.com 写道
> > >
> > > From: ***@hpe.com
> > > To: xcat-***@lists.sourceforge.net
> > > Cc:
> > > Date: 2017幎7月5日 䞊午12:37:37
> > > Subject: [xcat-user] xcat.genesis.doxcat: Unrecognized
> directive
> > (dest=)
> > >
> > > Hello,
> >
> > [...]
> >
> > --
> > Nicolas
> >
>
> --
> Nicolas
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> xCAT-user mailing list
> xCAT-***@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/xcat-user
>
>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>
>
>
> _______________________________________________
> xCAT-user mailing list
> xCAT-***@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/xcat-user
>

--
Nicolas Roosen
HPC Support System Engineer
Mob.: +33 777 161 256
Xiao Peng Wang
2017-07-06 15:04:17 UTC
Permalink
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
Nicolas Roosen
2017-07-06 15:32:51 UTC
Permalink
Sure, here it is:

Object name: n042
arch=x86_64
bmc=172.30.192.56
bmcpassword=admin
bmcport=0
bmcusername=admin
chain=runcmd=bmcsetup,osimage=centos7.3-x86_64-node-giga
cons=ipmi
groups=all,ipmi,node-giga
installnic=mac
ip=172.30.240.56
mac=0c:c4:7a:71:f6:ba
mgt=ipmi
netboot=xnba
nfsserver=172.30.240.1
postbootscripts=otherpkgs
postscripts=syslog,remoteshell,syncfiles
primarynic=mac
provmethod=centos7.3-x86_64-node-giga
serialflow=hard
serialport=1
serialspeed=115200
tftpdir=/tftpboot
tftpserver=172.30.240.1
xcatmaster=172.30.240.1

If that matters the nodes I'm installing are x86_64 Supermicro servers.
BIOS is set in "legacy" mode (*not* UEFI).

Cheers.

On 07/06/2017 05:04 PM, Xiao Peng Wang wrote:
> Looks like the 'n042' is not a free node. Could you show out the node
> definition by 'lsdef n042'? If you did not set hostip and bmcip for the
> node definition. Could you use the parameter [hostiprange=imageprofile]
> [bmciprange=bmciprange] for nodediscoverstart command?
>
> Best Regards
> ----------------------------------------------------------------------
> Wang Xiaopeng (王晓朋)
> Manager for HPC SW Dev: xCAT, ESSL, SMI, Test
> IBM China Systems Laboratory (CSL)
>

[...]

--
Nicolas
Xiao Peng Wang
2017-07-06 23:25:18 UTC
Permalink
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
Nicolas Roosen
2017-07-07 14:51:38 UTC
Permalink
On 07/07/2017 01:25 AM, Xiao Peng Wang wrote:
> for this n042 node that the mac address has been discovered
> 'mac=0c:c4:7a:71:f6:ba' so I guess the discovery process has been done
> or this node was used before? If it was a an old node, please clean up
> the mac attribute and run 'nodeset n042 offline'. Could you check
> whether the mac is the node which you are trying to discover?
> And could you check the target node has been added in 'discoverydata'
> table? by running command 'tabdump discoverydata'.
>

Haaa yes, you're right, excellent.

So my "no nodes free" problem was because every time I specified the
host MAC address in my "mkdef" command:

mkdef -t node -o spiro-n042-clu groups="all,ipmi,node-giga" arch=x86_64
mgt=ipmi bmc=172.30.192.56 mac=0c:c4:7a:71:f6:ba ip=172.30.240.56
provmethod="centos7.3-x86_64-node-giga"

As soon as I removed the MAC address, "nodediscoverstart" said:

nodediscoverstart -V noderange=n042
Sequential Discovery: Started:
Number of free node names: 1

====================Free Nodes===================
NODE HOST IP BMC IP
spiro-n042-clu 172.30.240.56 172.30.192.56


But still if I look at the node console, it fails:

Jul 7 09:18:05 172.30.241.71 (none) xcat.genesis.doxcat: Running
getdestiny --> 172.30.240.1:3001
Jul 7 09:18:25 172.30.241.71 (none) xcat.genesis.doxcat: Received destiny=
Jul 7 09:18:25 172.30.241.71 (none) xcat.genesis.doxcat: The destiny=,
destiny parameters=
Jul 7 09:18:25 172.30.241.71 (none) xcat.genesis.doxcat: Unrecognized
directive (dest=)
Jul 7 09:18:31 172.30.241.71 (none) xcat.genesis.doxcat: ... Will
retry xCAT in 90 seconds


I'll look into this next week.

Thanks.
Nicolas

> Best Regards
> ----------------------------------------------------------------------
> Wang Xiaopeng (王晓朋)
> Manager for HPC SW Dev: xCAT, ESSL, SMI, Test
> IBM China Systems Laboratory (CSL)
>
> Tel: 86-10-82453455
> Email: ***@cn.ibm.com
>
> ----- Original message -----
> From: Nicolas Roosen <***@hpe.com>
> To: <xcat-***@lists.sourceforge.net>
> Cc:
> Subject: Re: [xcat-user] 回复: xcat.genesis.doxcat: Unrecognized
> directive (dest=)
> Date: Thu, Jul 6, 2017 11:34 PM
> Sure, here it is:
>
> Object name: n042
> arch=x86_64
> bmc=172.30.192.56
> bmcpassword=admin
> bmcport=0
> bmcusername=admin
> chain=runcmd=bmcsetup,osimage=centos7.3-x86_64-node-giga
> cons=ipmi
> groups=all,ipmi,node-giga
> installnic=mac
> ip=172.30.240.56
> mac=0c:c4:7a:71:f6:ba
> mgt=ipmi
> netboot=xnba
> nfsserver=172.30.240.1
> postbootscripts=otherpkgs
> postscripts=syslog,remoteshell,syncfiles
> primarynic=mac
> provmethod=centos7.3-x86_64-node-giga
> serialflow=hard
> serialport=1
> serialspeed=115200
> tftpdir=/tftpboot
> tftpserver=172.30.240.1
> xcatmaster=172.30.240.1
>
> If that matters the nodes I'm installing are x86_64 Supermicro servers.
> BIOS is set in "legacy" mode (*not* UEFI).
>
> Cheers.
>
> On 07/06/2017 05:04 PM, Xiao Peng Wang wrote:
> > Looks like the 'n042' is not a free node. Could you show out the node
> > definition by 'lsdef n042'? If you did not set hostip and bmcip
> for the
> > node definition. Could you use the parameter
> [hostiprange=imageprofile]
> > [bmciprange=bmciprange] for nodediscoverstart command?
> >
> > Best Regards
> >
> ----------------------------------------------------------------------
> > Wang Xiaopeng (王晓朋)
> > Manager for HPC SW Dev: xCAT, ESSL, SMI, Test
> > IBM China Systems Laboratory (CSL)
> >
>
> [...]
>
> --
> Nicolas
Xiao Peng Wang
2017-07-08 00:02:56 UTC
Permalink
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
Loading...