Nut upsuser - May 2009 - New NUT user with HP R3000XR problem

Greetings,

I've just installed nut-2.4.1 on a Solaris 10 x86 machine to monitor a 
HP R3000XR UPS.

SunOS babylon4 5.10 Generic_138889-08 i86pc i386 i86pc Solaris


First of all, I encountered a single problem during the build.  After 
configuring, drivers/Makefile ends up containing the following line:

LIBNETSNMP_LDFLAGS = -R../lib -L/usr/sfw/lib -lnetsnmp -lgen [...]

This is clearly wrong, as runtime lib specifications cannot be relative.  
By correcting that to -R/usr/sfw/lib, everything compiled and installed 
fine.  

(This occurs regardless of whether or not snmp support is enabled.)


I've already seen from the list archives that the BCMXCP driver is the 
correct driver for this UPS.  I have already verified that I have the 
correct serial port and that it's working properly (I am able to 
communicate with an APC SU1400 on this port, using apcupsd-3.14.6 and an 
APC 940-0024C cable, with no problems).  I have set the port permissions 
as per the install instructions, I've set the baud rate on the port to 
9600 as per existing threads in the archives, and I'm using the HP 
straight-through serial cable (or is it just a serial extension cable?) 
that came with the UPS.  That cable bears the following identifying 
information:

HP 397237-002
SPARES P/N 397642-001 5006

ups.conf looks like this:

[tokamak]
    driver      = bcmxcp
    port        = /dev/tty00
    baud_rate   = 19200
    desc        = HP R3000XR (main rack)


This is where things start going pear-shaped.  I apparently cannot 
communicate with the UPS.  I can start the bcmxcp driver, but it fails 
to communicate at 9600, and "autodetects" a rate of 19200:

babylon4:root:/opt/nut/etc:9 # ../bin/upsdrvctl start
Network UPS Tools - UPS driver controller 2.4.1
Network UPS Tools - BCMXCP UPS driver 0.21 (2.4.1)
RS-232 communication subdriver 0.17
No response from UPS on /dev/tty00 with baudrate 9600
Attempting to autodect baudrate
Connected to UPS on /dev/tty00 with baudrate 19200

If I set baud_rate=19200 in ups.conf, that starts right up:

babylon4:root:/opt/nut/etc:11 # ../bin/upsdrvctl start
Network UPS Tools - UPS driver controller 2.4.1
Network UPS Tools - BCMXCP UPS driver 0.21 (2.4.1)
RS-232 communication subdriver 0.17
Connected to UPS on /dev/tty00 with baudrate 19200

But neither gets any model or version information from the UPS.  I can 
start upsd:

babylon4:root:/opt/nut/etc:18 # ../sbin/upsd
Network UPS Tools upsd 2.4.1
listening on 127.0.0.1 port 3493
listening on ::1 port 3493
/opt/nut/var is world readable
Connected to UPS [tokamak]: bcmxcp-tokamak
Maximum number of connections limited to 256 [requested 1024]

but can't get status from the UPS:

babylon4:root:/opt/nut/etc:19 # ../bin/upsc tokamak at localhost ups.status
Error: Connection failure: Connection refused


Am I using the correct cable here?  If not, what's the correct cable to 
use?

If I am already using the correct cable, I'm presuming this means my 
UPS's RS232 module is toast.



Thanks.


-- 
  Phil Stracchino, CDK#2     DoD#299792458     ICBM: 43.5607, -71.355
  alaric at caerllewys.net   alaric at metrocast.net   phil at co.ordinate.org
         Renaissance Man, Unix ronin, Perl hacker, Free Stater
                 It's not the years, it's the mileage.

On Tue, May 26, 2009 at 09:51:31PM -0400, Brother Railgun of Reason
wrote:
> ups.conf looks like this:
> 
> [tokamak]
>     driver      = bcmxcp
>     port        = /dev/tty00
>     baud_rate   = 19200
>     desc        = HP R3000XR (main rack)
(Oops.  That's *after* changing to the 19200 rate the driver claims to
autodetect.)

> Am I using the correct cable here?  If not, what's the correct cable to
> use?
The other detail I forgot to add was that, just out of sheer 
constructive paranoia, I tried using an APC 940-0024C cable instead of 
the HP cable, which as far as I can tell with a DVM is just a dumb 
serial cable wired all nine pins straight through.  With the HP cable, 
the driver thinks it has comm sync at 19200, but upsdrvctl can't talk to 
it; with the APC cable, the driver can't even establish communication at 
all.


-- 
  Phil Stracchino, CDK#2     DoD#299792458     ICBM: 43.5607, -71.355
  alaric at caerllewys.net   alaric at metrocast.net   phil at co.ordinate.org
         Renaissance Man, Unix ronin, Perl hacker, Free Stater
                 It's not the years, it's the mileage.

On Wed, May 27, 2009 at 08:42:40PM +0200, Kjell Claesson
wrote:
> > On Wed, May 27, 2009 at 06:04:40PM +0200, Kjell Claesson wrote:
> OK,
> 8<-----------------------------snip-----------------------------------
> > With the config indicating 19200, it locks immediately at 19200.  I
just
> > don't get any UPS "signature" displayed on connection:
> >
> > babylon4:root:/opt/nut:24 # bin/upsdrvctl start
> > Network UPS Tools - UPS driver controller 2.4.1
> > Network UPS Tools - BCMXCP UPS driver 0.21 (2.4.1)
> > RS-232 communication subdriver 0.17
> > Connected to UPS on /dev/tty00 with baudrate 19200
> This is right. You dont get anything more from the driver.
Oh, OK ...  I thought the examples and other list threads with R3000XRs 
showed a version string being echoed at this point.

> 
> > babylon4:root:/opt/nut:25 # sbin/upsd
> > Network UPS Tools upsd 2.4.1
> > listening on 127.0.0.1 port 3493
> > listening on ::1 port 3493
> > /opt/nut/var is world readable
> > Connected to UPS [tokamak]: bcmxcp-tokamak
> > Maximum number of connections limited to 256 [requested 1024]
> > babylon4:root:/opt/nut:26 # bin/upsc tokamak at localhost
> > Error: Connection failure: Connection refused
> This is not so good. But it is the daemon that is complaining
> that it can not connect to the socket right.
> 
> Check your configuration once more.
I've been over it several times ...  it appears upsd is dying 
immediately after startup, but I don't yet know why.

> You have a config block. (not used at the moment.
[...]
> Yes it looks like it is giving the right sequence for load-segments.
So far, so good then...
> So check that all permissions on the files and socket directory is right.
> The driver is working OK.
The way I configured it, that should be /opt/nut/var:

[from config.status:]

  exec /bin/bash "./configure" '--prefix=/opt/nut' 
'--htmldir=/var/httpd/local/manuals/nut' '--with-serial' 
'--with-gd-includes=-I/usr/local/include' '--with-gnu-ld' 
'--with-statepath=/opt/nut/var' '--with-pidpath=/opt/nut/var' 
'--with-user=nut' '--with-group=nut' '--with-cgi'
'--with-lib'
 '--without-usb' '--without-snmp' '--without-ssl' 
$ac_configure_extra_args --no-create --no-recursion


babylon4:root:/opt/nut:53 # ls -l . var
.:
total 15
drwxr-xr-x 2 root nut 44 May 26 21:02 bin/
drwxr-xr-x 2 root nut  5 May 26 21:02 cgi-bin/
drwxr-x--- 2 root nut 15 May 26 21:15 etc/
drwxr-xr-x 2 root nut  6 May 26 21:02 html/
drwxr-xr-x 2 root nut  2 May 26 21:02 include/
drwxr-xr-x 3 root nut  7 May 26 21:02 lib/
drwxr-xr-x 2 root nut  5 May 26 21:02 sbin/
drwxr-xr-x 3 root nut  5 May 26 21:02 share/
drwxrwx--- 2 root nut  4 May 27 16:51 var/

var:
total 2
srw-rw---- 1 nut nut 0 May 27 16:51 bcmxcp-tokamak-rw-r--r-- 1 nut nut 4 May 27
16:51 bcmxcp-tokamak.pid

babylon4:root:/opt/nut:54 # ls -l /dev/tty00
lrwxrwxrwx 1 root root 26 Apr 18 18:48 /dev/tty00 -> 
../devices/isa/asy at 1,3f8:a
babylon4:root:/opt/nut:55 # ls -l /devices/isa/asy at 1,3f8:a
crw-rw---- 1 root nut 106, 0 May 27 17:17 /devices/isa/asy at 1,3f8:a



-- 
  Phil Stracchino, CDK#2     DoD#299792458     ICBM: 43.5607, -71.355
  alaric at caerllewys.net   alaric at metrocast.net   phil at co.ordinate.org
         Renaissance Man, Unix ronin, Perl hacker, Free Stater
                 It's not the years, it's the mileage.