freeipmi-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Freeipmi-users] Node "not discovered" using ipmipower


From: Al Chu
Subject: Re: [Freeipmi-users] Node "not discovered" using ipmipower
Date: Fri, 12 Apr 2024 10:14:55 -0700
User-agent: Mozilla Thunderbird

Hi Dave,

That is quite interesting.  Short description, in the background of shell mode, ipmipower will "ping" nodes half regularly to make sure they are there.  Because it is supposed to "run forever" under powerman, there's no point to send power control to nodes missing from a cluster (e.g. taken out for maintenance) and thus slowing down powerman.

"not discovered" means that it can't ping that node.

Does "rmcpping" to the node work?

if not, does "ipmiping" to the node work?

If so, perhaps the "ipmiping" workaround to ipmipower would work (see manpage on workaround).

Al

On 4/11/24 21:21, Devon Bautista via Freeipmi-users wrote:
Hello,

Similar to my previous email regarding issues with the M50CYP Intel board, I'm running into issues with the newer M50FCP Intel board.

I'm trying to use powerman to manage power to nodes with an Intel M50FCP board:

[admin@te-head ~]$ pm -q
on:
off:
unknown: te01

The /etc/powerman/powerman.conf is:

include "/etc/powerman/ipmipower.dev"
device  "ipmi0"  "ipmipower"  "/usr/sbin/ipmipower -D lanplus -u admin -p <password> -h bmc-te01 -I 17 |&"
node "te01" "ipmi0" "bmc-te01"

Running the ipmipower command:

[admin@te-head ~]$ sudo /usr/sbin/ipmipower -D lanplus -u admin -p <password> -h -I 17 bmc-te01
ipmipower> debug
debugging is now on
ipmipower> stat
bmc-te01: not discovered

However:

[admin@te-head ~]$ sudo /usr/sbin/ipmipower -D lanplus -u admin -p <password> -h bmc-te01 -I 17 --stat
bmc-te01: on

I'm also able to hit the BMC with ipmitool:

[admin@te-head ~]$ ipmitool -I lanplus -H bmc-te01 -U admin -P <password> power status
Chassis Power is on

The "not discovered" error doesn't seem to be documented in the ipmipower manual page and I haven't been able to find out online what it means. It is odd that entering "stat" in the ipmipower shell errs, but using the "--stat" option works fine. This only seems to happen with these newer Intel Fox Creek Pass boards and not other boards I have.

For context, conman works with the following config:

SERVER keepalive=ON
SERVER logdir="/var/log/conman"
SERVER logfile="/var/log/conman.log"
SERVER loopback=ON
SERVER pidfile="/var/run/conman.pid"
SERVER resetcmd="/usr/bin/powerman -0 %N; sleep 5; /usr/bin/powerman -1 %N"
SERVER tcpwrappers=ON
GLOBAL seropts="115200,8n1"
GLOBAL log="/var/log/conman/console.%N"
GLOBAL logopts="sanitize,timestamp"# IPMI Console\
CONSOLE name="te01" dev="ipmi:bmc-te01" ipmiopts="U:admin,P:<password>,C:17,W:solpayloadsize"

Is there a way to fix this?

Versions:
ipmipower - 1.6.8
ipmiconsole - 1.6.8
powerman - 2.3.26

Regards,
Devon

_______________________________________________
Freeipmi-users mailing list
Freeipmi-users@gnu.org
https://lists.gnu.org/mailman/listinfo/freeipmi-users

--
Al Chu
Livermore Computing
Lawrence Livermore National Laboratory




reply via email to

[Prev in Thread] Current Thread [Next in Thread]