Re: [Freeipmi-devel] Re: FreeIPMI - Alpha-4 QA-1 Released.

[Joseph Ruscio]

Albert,

Does this happen on more than one node? I just set up a script to flush the 
cache and run sensors 200 times (takes about 2 1/2 hours) and it did not fail 
to cache all 107 nodes a single time.

thanks,
Joe

On Friday 12 March 2004 06:49 pm, Albert Chu wrote:
> There seems to be a bug in the sensors code.  The code does not always
> consistently cache all 107 sensors on the tiger4s.  It sometimes finds
> less than that, and subsequently can't read those sensors.  Its happened
> to me 4 times out of maybe 30-35 tries.
>
> here's one instance
>
> tdev3|/home/achur 148>/usr/sbin/sensors -f
> flushing cache... done
> tdev3|/home/achur 149>time /usr/sbin/sensors
> Creating sensors data repository cache. This may take a while...
> 1: I/O Bd +1.3V (Voltage): N/A (low=1.23/nom=1.30/high=1.36) [FAILED]
> 2: I/O Bd +1.5V (Voltage): 1.53 V (low=1.42/nom=1.50/high=1.57) [OK]
> 3: I/O Bd +1.8V (Voltage): 1.84 V (low=1.70/nom=1.80/high=1.88) [OK]
> 4: I/O Bd +3.3V SB (Voltage): 3.25 V (low=3.13/nom=3.30/high=3.45) [OK]
> 5: I/O Bd +3.3V_A (Voltage): 3.40 V (low=3.06/nom=3.30/high=3.52) [OK]
> 6: I/O Bd +3.3V_B (Voltage): 3.38 V (low=3.12/nom=3.30/high=3.46) [OK]
> 7: IORISER CMOS Bat (Voltage): 3.25 V (low=2.40/nom=2.99/high=3.60) [OK]
> 8: I/O Bd +5V_A (Voltage): 5.02 V (low=4.73/nom=4.99/high=5.23) [OK]
> 9: I/O Bd +5V_B (Voltage): 5.03 V (low=4.74/nom=5.00/high=5.23) [OK]
> 10: I/O Bd +12V (Voltage): 11.84 V (low=11.28/nom=11.90/high=12.46) [OK]
> 11: I/O Bd -12V (Voltage): -11.92 V (low=-12.64/nom=-12.06/high=-11.49)
> [OK] 12: Proc Bd +1.2V (Voltage): 1.22 V (low=1.18/nom=1.20/high=1.21)
> [ALERT] 13: MEM Bd 1 +1.25V (Voltage): 1.28 V (low=1.18/nom=1.25/high=1.31)
> [OK] 14: MEM Bd 2 +1.25V (Voltage): 1.28 V (low=1.18/nom=1.25/high=1.31)
> [OK] 15: Proc Bd +1.3V (Voltage): 1.32 V (low=1.22/nom=1.30/high=1.36) [OK]
> 16: Proc Bd +1.5V (Voltage): 1.52 V (low=1.45/nom=1.50/high=1.54) [OK] 17:
> Proc Bd +1.8V (Voltage): 1.84 V (low=1.73/nom=1.80/high=1.87) [OK] 18: Proc
> Bd +3.3V SB (Voltage): 3.25 V (low=3.13/nom=3.29/high=3.46) [OK] 19: Proc
> Bd +3.3V (Voltage): 3.30 V (low=3.13/nom=3.29/high=3.46) [OK] 20: IORISER
> +12V SB (Voltage): 12.45 V (low=11.21/nom=11.80/high=12.39) [ALERT]
> 21: IORISER +2.5V (Voltage): 2.51 V (low=2.35/nom=2.47/high=2.60) [OK]
> 22: IORISER +1.5V SB (Voltage): 1.52 V (low=1.42/nom=1.50/high=1.57) [OK]
> 23: IORISER +1.5V (Voltage): 1.52 V (low=1.42/nom=1.50/high=1.57) [OK]
> 24: IORISER +5V SB (Voltage): 4.93 V (low=4.67/nom=4.93/high=5.16) [OK]
> 25: I/O Bd SCSI +5V (Voltage): 5.06 V (low=4.63/nom=4.99/high=5.33) [OK]
> 26: Mem Bd 1 Temp (Temperature): 30.00 C
> (low=10.00/nom=33.00/high=45.00) [OK]
> 27: Mem Bd 2 Temp (Temperature): 28.00 C
> (low=10.00/nom=33.00/high=45.00) [OK]
> 28: I/O Bd Temp 1 (Temperature): 34.00 C
> (low=10.00/nom=46.00/high=54.00) [OK]
> 29: I/O Bd SIOH Temp (Temperature): 37.00 C
> (low=10.00/nom=55.00/high=95.00) [OK]
> 30: I/O Bd Temp 3 (Temperature): 33.00 C
> (low=10.00/nom=49.00/high=57.00) [OK]
> 31: Proc Bd Amb Temp (Temperature): 23.00 C
> (low=10.00/nom=25.00/high=38.00) [OK]
> 32: Proc Bd SNC Temp (Temperature): 42.00 C
> (low=10.00/nom=55.00/high=95.00) [OK]
> 33: F38 Tach Fan 1 (Fan): 2620.00 RPM
> (low=2260.00/nom=2600.00/high=2880.00) [OK]
> 0.187u 0.002s 0:11.44 1.5%    0+0k 0+0io 259pf+0w
> tdev3|/home/achur 150>time /usr/sbin/sensors -s 34
> sensors: error: Current reading not available for sensors [34]. You may
> try "-vv" option for more info
> 0.184u 0.003s 0:00.19 94.7%   0+0k 0+0io 246pf+0w
>
> --
> Albert Chu
> address@hidden
> Lawrence Livermore National Laboratory
>
> ----- Original Message -----
> From: address@hidden
> Date: Friday, March 12, 2004 1:39 pm
> Subject: Re: FreeIPMI - Alpha-4 QA-1 Released.
>
> > Hi Al,
> > Currently it can be done using
> > # fish --sms-io-base=VALUE --script-file=SCRIPT-PATH
> > But it makes sense to support a config-file option and
> > command-line argument.
> >
> > Alpha5:
> > * New feature requests - closing data: Monday 15 Mar 2004 9:00PM.
> > * Alpha5 QA0 - release - Tuesday 16 Mar 2004 (estimated).
> > * Alpha5 - release - Thursday 18 Mar 2004 (estimated).
> >
> > Robin, Joe,
> > Should the tech-specs reflect these new features and get approved
> > again ?
> >
> > Happy Hacking,
> > --ab
> >
> > > Hey AB,
> > >
> > > I just realized this.  For sensors and bmc-config, I think we should
> > > also have a --io-port option.  So they can be used on machine sthat
> > > don't use 0xCA2 as their KCS i/o port ...
> > >
> > > Al
> > >
> > > --
> > > Albert Chu
> > > address@hidden
> > > Lawrence Livermore National Laboratory
> > >
> > > ----- Original Message -----
> > > From: Albert Chu <address@hidden>
> > > Date: Friday, March 12, 2004 9:48 am
> > > Subject: Re: FreeIPMI - Alpha-4 QA-1 Released.
> > >
> > >> Hey AB,
> > >>
> > >> A few issues with the sensors and bmc-config stuff.
> > >>
> > >> sensors
> > >>
> > >> It seems the sensors cache is always stored in /var/tmp.  I don't
> > >> thinkthis is the greatest location for it.  We have cronjobs that
> > >> regularlycleanup/remove files in /var/tmp.  Perhaps we could have
> > >> an option that
> > >> allows us to specify the location of the sensors cache?  Or
> >
> > perhaps a
> >
> > >> fish configuratoin file can always be read which lists an alternate
> > >> default location.  It might be best to ask Robin and the admin team
> > >> whatthe best option is though.
> > >>
> > >> bmc-config
> > >>
> > >> I assume that some of the issues we talked about in the last e-mail
> > >> thread about bmc-config just aren't done yet, so I won't talk about
> > >> them.
> > >>
> > >> I did notice a problem with userX_password configuration in the
> > >> configuration file.  If you don't list a password for an already
> > >> existent user, the password is automatically changed to NULL.  That
> > >> means we'd always have to re-type the password in the configuration
> > >> fileif we want to re-commit that configuration file.
> > >>
> > >> How about if userX_password is a special case and is always
> >
> > commented>> out when you do a --checkout?  That way, the values are
> > only>> changed if
> >
> > >> the user really wants them changed. So ...
> > >>
> > >> ###   Uncomment following to assign password
> > >> ###   Empty password is acceptable.
> > >> #user1_password
> > >> #user2_password
> > >> #user3_password
> > >> #user4_password
> > >>
> > >> Al
> > >>
> > >> --
> > >> Albert Chu
> > >> address@hidden
> > >> Lawrence Livermore National Laboratory
> > >>
> > >> ----- Original Message -----
> > >> From: Anand Babu <address@hidden>
> > >> Date: Wednesday, March 10, 2004 6:11 am
> > >> Subject: FreeIPMI - Alpha-4 QA-1 Released.
> > >>
> > >> >     ********* THIS IS NOT A RELEASE. ONLY FOR QA TEAM *********
> > >> > Alpha-4 QA-1 Released:
> > >> >
> > >> > ftp://ftp.californiadigital.com/pub/freeipmi/fish-0.1.0-alpha4-
> >
> > qa1.tar.gzftp://ftp.californiadigital.com/pub/freeipmi/libfreeipmi-
> >
> > >> > 0.0.0-alpha4-qa1.tar.gz
> > >> >
> > >> >     ********* THIS IS NOT A RELEASE. ONLY FOR QA TEAM *********
> > >> >
> > >> > After installing of "fish", you will find "sensors" and "bmc-
> >
> > config">> > installed in the ${prefix}/sbin dir. Man pages are
> > available for
> >
> > >> both> the utilities.
> > >>
> > >> > Whats New!
> > >> > ----------
> > >> >  - Closed all reported BUGS [#8062 #8064 #8067 #8075 #8076 #8082
> > >> > #8084].  - Proper exit status.
> > >> >
> > >> > Happy Hacking,
> > >> > --
> > >> > _.|_
> > >> > (_||_)
> > >> > Free as in Freedom <www.gnu.org>
>
> _______________________________________________
> Freeipmi-devel mailing list
> address@hidden
> http://mail.nongnu.org/mailman/listinfo/freeipmi-devel

-- 
******************************************
Joseph F. Ruscio
PhD Candidate
Computing Systems Research Lab
Virginia Tech
******************************************