Monitoring Solaris 10 and OpenSolaris

[lyle]

It's interesting to hear about Solaris monitoring projects that may be in the works. But I *need* an nrpe client working on Solaris (V9 and 10 for Sparc).

Last time I tried compiling from the nrpe source on Nagios Exchange, I got complaints about missing SSL packages. When I tried to run the compiled nrpe, it just exited silently. Tried running a "truss" on the executable gave me information, but nothing obvious to my pea brain.

I know Nagios Corp doesn't have Solaris expertise on the team, but I need to get nrpe running somehow. Our corporate databases are all Oracle on Solaris. Not an uncommon config, I'd think.

The best article I've found on the net is: http://www.utahsysadmin.com/2008/03/14/ ... olaris-10/ although it's x86 oriented.

Is something like this best handled as an email support request, rather than on the forum?

Thanks...Lyle

[tonyyarusso]

Lyle, you could try looking into any of these options for now:

1. Builds by Op5
2. Instructions pertaining to restricted ciphers
3. Using blastwave.org packages

[lyle]

Trying to get nrpe and the plugins installed on Solaris is going a bit slow. I can't even install the plugins from source.

On a Sparc Solaris 10 machine, when trying to install nagios-plugins-1.4.15 configure runs fine, but make dies with:

Code: Select all

/bin/bash ../libtool --tag=CC    --mode=link gcc -DNP_VERSION='"1.4.15"' -g -O2   -L. -L/usr/local/ssl/lib -o check_ldap check_ldap.o netutils.o utils.o ../lib/libnagiosplug.a ../gl/libgnu.a  -lnsl -lsocket -lresolv -lldap -llber -lintl -lpthread -ldl
libtool: link: gcc -DNP_VERSION=\"1.4.15\" -g -O2 -o check_ldap check_ldap.o netutils.o utils.o  -L/root/nagios-plugins-1.4.15/plugins -L/usr/local/ssl/lib ../lib/libnagiosplug.a ../gl/libgnu.a /usr/local/lib/libldap.so -L/usr/local/pgsql/lib -L/usr/local/lib -L/usr/openwin/lib -L/usr/lib -L/usr/local/BerkeleyDB.4.2/lib -L/usr/X11R6/lib /usr/local/lib/libsasl2.so -L/usr/openwinlib -lssl -lcrypto /usr/local/lib/liblber.so -lresolv -lgen -lnsl -lnet -lsocket /usr/local/lib/libintl.so /usr/local/lib/libiconv.so -lsec -lc -lpthread -ldl -R/usr/local/lib -R/usr/local/lib
ld: fatal: library -lnet: not found
ld: fatal: File processing errors. No output written to check_ldap
collect2: ld returned 1 exit status
*** Error code 1
make: Fatal error: Command failed for target `check_ldap'
Current working directory /root/nagios-plugins-1.4.15/plugins
*** Error code 1
The following command caused the error:
failcom='exit 1'; \
for f in x $MAKEFLAGS; do \
  case $f in \
    *=* | --[!k]*);; \
    *k*) failcom='fail=yes';; \
  esac; \
done; \
dot_seen=no; \
target=`echo all-recursive | sed s/-recursive//`; \
list='gl tap lib plugins plugins-scripts plugins-root po'; for subdir in $list; do \
  echo "Making $target in $subdir"; \
  if test "$subdir" = "."; then \
    dot_seen=yes; \
    local_target="$target-am"; \
  else \
    local_target="$target"; \
  fi; \
  (cd $subdir && make  $local_target) \
  || eval $failcom; \
done; \
if test "$dot_seen" = "no"; then \
  make  "$target-am" || exit 1; \
fi; test -z "$fail"
make: Fatal error: Command failed for target `all-recursive'
Current working directory /root/nagios-plugins-1.4.15
*** Error code 1
make: Fatal error: Command failed for target `all'

I suspect there's some problem with the openssl distributed with Solaris, but I really have no clue.

Should I address this to [email protected]?

Thanks for any advice...Lyle

[tonyyarusso]

My point #2 about restricted ciphers is probably the issue with SSL you're running into, but yes, the community mailing lists would almost certainly be better equipped to walk you through this one.

[lyle]

Thanks for the response, Tony. I read what you're referring to a little closer this time. I'll post the plugin question to the nagiosplug-help list.

Switching topics to the nrpe daemon (2.12), I used the suggested switches on the configure command and the logging mods to nrpe.c
Since my system is the 08/07 release (Solaris 10 8/07 s10s_u4wos_12b SPARC), it has the SUNWcry and SUNWcryr packages, so I didn't use the SSL_CTX_set_cipher_list patch.
I didn't notice any complaints during the whole configure/make/install procedure, and nrpe seemed to start ok as a SMF service.

Oddly, when I try to run check_nrpe either from the localhost that runs the daemon, or externally from my Nagios server, I get:

ld.so.1: check_nrpe: fatal: libssl.so.0.9.7: open failed: No such file or directory

Any advice or thoughts are appreciated....Lyle

[jtata]

My previous postion was an almost all Solaris environment and we monitored extensively with Nagios core using a customized check_by_ssh plugin. IIRC we had CPU, Mem, Disk usage on key partitions and output from PRTDIAG fed to a perl/expect script to alarm on errors. We did not experience much of an issue with multiple connections, I belive checking each service every 10 minutes on ~200 hosts.

[rdedon]

Lyle,
please update with results if you are still having issues after coordinating with nagiosplug-help.

Thanks!

[lyle]

Well I think I finally got things running, on a test client at least.

The best resources continued to be:
http://www.utahsysadmin.com/2008/03/14/ ... olaris-10/
though there was a lot to wade through.

My client host is Solaris 10 Sparc 8/07 (cat /etc/release to see). 8/07 comes with SUNWcry and SUNWcryr, the 256bit encryption library (see FAQ F0372), so I decided to use them.

For the plugins:
The plugins mail list ([email protected]) did come back with one response, saying it looked like the make couldn't find "libnet.so". The error I listed earlier seemed to have a problem with "lnet", but I wouldn't have guessed that was an entry point in libnet.so
I did a find for the libnet.so, decided to try the 32 bit version, so added "/usr/j2se/jre/lib/sparc" to the end of my LD_LIBRARY_PATH
After that, make stopped complaining, and a "make install" moved the plugins to libexec.
They seem to work.

For the nrpe daemon:
Following instructions on the site mentioned, I did:

Code: Select all

configure --with-ssl=/usr/sfw --with-ssl-lib=/usr/sfw/lib

Again, per the instructions, I modified nrpe.c lines 616-619 to deal with the logfile issue, after which, make and make install worked.
Installation with SMF worked fine, following the instructions.
But running check_nrpe locally died, saying "libssl.so.0.9.7 open failed"
At the end of the article (post 93), it points out that a problem with the makefile meant the correct RPATH wasn't compiled into the binary.
You can see the problem via "ldd check_nrpe"
I recompiled via:

Code: Select all

LDFLAGS=-R/usr/local/ssl/lib ./configure --with-ssl=/usr/sfw --with-ssl-lib=/usr/sfw/lib

.
make, make install, stop & start the daemon, and by golly....it worked. I can query the daemon both locally and from the Nagios server. Hope my luck holds out.

To me, Solaris is still very much in the main stream. I hope that XI will soon come with a productized client like the one for Centos.

...Lyle

[tonyyarusso]

Glad to hear it worked out. As far as future plans, I do not expect the XI server to ever run on Solaris, but we should have some kind of client coming, yes.

[lyle]

Any news on a packaged client for Solaris?

I'm back trying to get nrpe and plugins installed on Solaris 9 & 10 (sparc) and having problems again.

Thanks....Lyle