Xymon Mailing List Archive search

Preliminary patch for RC3 brokenness

15 messages in this thread

list Henrik Størner · Thu, 24 Feb 2005 00:19:09 +0100 ·
is available at http://www.hswn.dk/beta/post-RC3.patch

It's been running for an hour now without problems, so I think I've
got the embarassing bugs fixed:

* hobbitd_larrd should not crash anymore
* the alert-configuration should load correctly
* runtime-libraries should hopefully be found correctly on Linux and *BSD
* maint.pl should no longer complain about un-initialised variables

In addition, you get:
* A new hobbit-mailack tool was added, so you can acknowledge 
  alerts via e-mail without having to use BB's bb-mailack.sh
  (see the hobbit-mailack man-page for instructions on setting it up)
* Support for OpenBSD was added

To install the patch, do:

   cd hobbit-4.0-RC3
   patch -p0 </tmp/post-RC3.patch
   rm Makefile
   ./configure
   make
   su -c "make install"

Note that you MUST delete the old Makefile, and re-run the configure
script.


Regards,
Henrik
list Bruce Lysik · Wed, 23 Feb 2005 15:54:11 -0800 ·
Hmm.  I just got some odd page with RC3-patch.

This group:

HOST=$HG-IMAGE
       SCRIPT /opt/bb/server/ext/email onduty SERVICE=procs RECOVERED        SCRIPT /opt/bb/server/ext/email ops    SERVICE=procs RECOVERED        SCRIPT /opt/bb/server/ext/email onduty SERVICE=http,conn DURATION>6 RECOVERED        SCRIPT /opt/bb/server/ext/email ops    SERVICE=http,conn DURATION>6 RECOVERED        SCRIPT /opt/bb/server/ext/email ops    EXSERVICE=procs,http,conn RECOVERED 
And I just got some cpu alerts sent to the onduty address.  In fact, the 'info' page looks totally wrong for alerts.    Ack.  Just got another alert I shouldn't have.  (A purple for a vmstat monitor which should have only gone to the 'ops' address.)

--
Bruce Z. Lysik  <user-4e63a10f8934@xymon.invalid>
Operations Engineer
quoted from Henrik Størner

-----Original Message-----
From: Henrik Stoerner [mailto:user-ce4a2c883f75@xymon.invalid]
Sent: Wednesday, February 23, 2005 3:19 PM
To: user-ae9b8668bcde@xymon.invalid
Subject: [hobbit] Preliminary patch for RC3 brokenness


is available at http://www.hswn.dk/beta/post-RC3.patch

It's been running for an hour now without problems, so I think I've
got the embarassing bugs fixed:

* hobbitd_larrd should not crash anymore
* the alert-configuration should load correctly
* runtime-libraries should hopefully be found correctly on Linux and *BSD
* maint.pl should no longer complain about un-initialised variables

In addition, you get:
* A new hobbit-mailack tool was added, so you can acknowledge   alerts via e-mail without having to use BB's bb-mailack.sh
  (see the hobbit-mailack man-page for instructions on setting it up)
* Support for OpenBSD was added

To install the patch, do:

   cd hobbit-4.0-RC3
   patch -p0 </tmp/post-RC3.patch
   rm Makefile
   ./configure
   make
   su -c "make install"

Note that you MUST delete the old Makefile, and re-run the configure
script.


Regards,
Henrik

list Bruce Lysik · Wed, 23 Feb 2005 15:59:16 -0800 ·
And I just got some cpu alerts sent to the onduty address.  In fact, the 'info' page looks totally wrong for alerts.    Ack.  Just got another alert I shouldn't have.  (A purple for a vmstat monitor which should have only gone to the 'ops' address.)
I've confirmed that this is an issue with the patch.  I rolled back to RC3 without patch, and suddenly the alerting definitions that are displayed on the 'info' pages make sense again.

--
Bruce Z. Lysik  <user-4e63a10f8934@xymon.invalid>
Operations Engineer
list Andy France · Thu, 24 Feb 2005 15:02:44 +1300 ·

Hi Henrik,

For my Solaris x86 Hobbit server, I use the Blastwave packages
(http://www.blastwave.org) for installing the libraries required for
compiliation.

Would it be possible to get the following search directories added to the
build scripts to use the Blastwave /opt/csw tree as well?

build/fping.sh:    for DIR in /opt/csw/sbin ...
build/ldap.sh:     for DIR in /opt/csw ...
build/pcre.sh:     for DIR in /opt/csw ...
build/rrd.sh:      for DIR in /opt/csw ...
build/ssl.sh:      for DIR in /opt/csw/ssl ...

Up to now I've been editing these scripts before building each release.

Is anybody else using this platform and package combination?

Regards,
Andy.

#####################################################################################

This email is intended for the person to whom it is addressed
only. If you are not the intended recipient, do not read, copy
or use the contents in any way. The opinions expressed may not
necessarily reflect those of ZESPRI Group of Companies ('ZESPRI').

While every effort has been made to verify the information
contained herein, ZESPRI does not make any representations 
as to the accuracy of the information or to the performance
of any data, information or the products mentioned herein.
ZESPRI will not accept liability for any losses, damage or
consequence, however, resulting directly or indirectly from
the use of this e-mail/attachments.
#####################################################################################
list Asif Iqbal · Wed, 23 Feb 2005 21:24:30 -0500 ·
quoted from Bruce Lysik
On Thu, Feb 24, 2005 at 12:19:09AM, Henrik Stoerner wrote:
is available at http://www.hswn.dk/beta/post-RC3.patch

It's been running for an hour now without problems, so I think I've
got the embarassing bugs fixed:

* hobbitd_larrd should not crash anymore
* the alert-configuration should load correctly
* runtime-libraries should hopefully be found correctly on Linux and *BSD
* maint.pl should no longer complain about un-initialised variables

In addition, you get:
* A new hobbit-mailack tool was added, so you can acknowledge 
  alerts via e-mail without having to use BB's bb-mailack.sh
  (see the hobbit-mailack man-page for instructions on setting it up)
* Support for OpenBSD was added

To install the patch, do:

   cd hobbit-4.0-RC3
   patch -p0 </tmp/post-RC3.patch
   rm Makefile
   ./configure
   make
   su -c "make install"
May be another silly error :-) . This is what I get while trying to
compile

gcc -g -O2 -Wall -Wno-unused -D_REENTRANT -DHAVE_RPCENT -DMAXMSG=32768
-DBBDPORTNUMBER=1984 -I. -I/usr/share/src/hobbit-4.0-RC3/include -o
hobbit-mailack  hobbit-mailack.o ../lib/libbbgen.a -L/usr/local/lib
-lpcre
Undefined                       first referenced
 symbol                             in file
recv                                ../lib/libbbgen.a(sendmsg.o)
getservbyname                       ../lib/libbbgen.a(sendmsg.o)
gethostbyname                       ../lib/libbbgen.a(sendmsg.o)
socket                              ../lib/libbbgen.a(sendmsg.o)
getsockopt                          ../lib/libbbgen.a(sendmsg.o)
connect                             ../lib/libbbgen.a(sendmsg.o)
inet_aton                           ../lib/libbbgen.a(sendmsg.o)
inet_ntoa                           ../lib/libbbgen.a(sendmsg.o)
shutdown                            ../lib/libbbgen.a(sendmsg.o)
ld: fatal: Symbol referencing errors. No output written to
hobbit-mailack
collect2: ld returned 1 exit status
make[1]: *** [hobbit-mailack] Error 1
make[1]: Leaving directory `/usr/share/src/hobbit-4.0-RC3/hobbitd'
make: *** [hobbitd-build] Error 2


Thanks
quoted from Bruce Lysik
Note that you MUST delete the old Makefile, and re-run the configure
script.


Regards,
Henrik

-- 

Asif Iqbal
PGP Key: 0xE62693C5 KeyServer: pgp.mit.edu
"It is not the strongest of the species that survives, not the most intelligent, but
 the one most responsive to change."    - Charles Darwin
list Andy France · Thu, 24 Feb 2005 16:42:12 +1300 ·

Asif Iqbal <user-c8222abeff59@xymon.invalid> wrote on 24/02/2005 15:24:30:
quoted from Asif Iqbal
On Thu, Feb 24, 2005 at 12:19:09AM, Henrik Stoerner wrote:
is available at http://www.hswn.dk/beta/post-RC3.patch
May be another silly error :-) . This is what I get while trying to
compile
gcc -g -O2 -Wall -Wno-unused -D_REENTRANT -DHAVE_RPCENT -DMAXMSG=32768
-DBBDPORTNUMBER=1984 -I. -I/usr/share/src/hobbit-4.0-RC3/include -o
hobbit-mailack  hobbit-mailack.o ../lib/libbbgen.a -L/usr/local/lib
-lpcre
Undefined                       first referenced
symbol                             in file
recv                                ../lib/libbbgen.a(sendmsg.o)
getservbyname                       ../lib/libbbgen.a(sendmsg.o)
gethostbyname                       ../lib/libbbgen.a(sendmsg.o)
socket                              ../lib/libbbgen.a(sendmsg.o)
getsockopt                          ../lib/libbbgen.a(sendmsg.o)
connect                             ../lib/libbbgen.a(sendmsg.o)
inet_aton                           ../lib/libbbgen.a(sendmsg.o)
inet_ntoa                           ../lib/libbbgen.a(sendmsg.o)
shutdown                            ../lib/libbbgen.a(sendmsg.o)
ld: fatal: Symbol referencing errors. No output written to
hobbit-mailack
collect2: ld returned 1 exit status
make[1]: *** [hobbit-mailack] Error 1
make[1]: Leaving directory `/usr/share/src/hobbit-4.0-RC3/hobbitd'
make: *** [hobbitd-build] Error 2
Thanks
--
Asif Iqbal
PGP Key: 0xE62693C5 KeyServer: pgp.mit.edu
"It is not the strongest of the species that survives, not the most
intelligent, but
the one most responsive to change."    - Charles Darwin
Me too :-)


gcc -g -O2 -Wall -Wno-unused -D_REENTRANT -DHAVE_RPCENT -DMAXMSG=32768
-DBBDPORTNUMBER=1984 -I. -I/export/home/hobbit/src/hobbit-4.0-RC3/include
-o hobbit-mailack  hobbit-mailack.o ../lib/libbbgen.a -L/opt/csw/lib -lpcre
quoted from Asif Iqbal
Undefined                       first referenced
 symbol                             in file
recv                                ../lib/libbbgen.a(sendmsg.o)
getservbyname                       ../lib/libbbgen.a(sendmsg.o)
gethostbyname                       ../lib/libbbgen.a(sendmsg.o)
socket                              ../lib/libbbgen.a(sendmsg.o)
getsockopt                          ../lib/libbbgen.a(sendmsg.o)
connect                             ../lib/libbbgen.a(sendmsg.o)
inet_aton                           ../lib/libbbgen.a(sendmsg.o)
inet_ntoa                           ../lib/libbbgen.a(sendmsg.o)
shutdown                            ../lib/libbbgen.a(sendmsg.o)
ld: fatal: Symbol referencing errors. No output written to hobbit-mailack
collect2: ld returned 1 exit status

gmake[1]: *** [hobbit-mailack] Error 1
gmake[1]: Leaving directory
`/export/home/hobbit/src/hobbit-4.0-RC3/hobbitd'
gmake: *** [hobbitd-build] Error 2
quoted from Andy France


Andy.

#####################################################################################

This email is intended for the person to whom it is addressed
only. If you are not the intended recipient, do not read, copy
or use the contents in any way. The opinions expressed may not
necessarily reflect those of ZESPRI Group of Companies ('ZESPRI').

While every effort has been made to verify the information
contained herein, ZESPRI does not make any representations 
as to the accuracy of the information or to the performance
of any data, information or the products mentioned herein.
ZESPRI will not accept liability for any losses, damage or
consequence, however, resulting directly or indirectly from
the use of this e-mail/attachments.
#####################################################################################
list Asif Iqbal · Wed, 23 Feb 2005 23:48:51 -0500 ·
quoted from Asif Iqbal
On Wed, Feb 23, 2005 at 09:24:30PM, Asif Iqbal wrote:
On Thu, Feb 24, 2005 at 12:19:09AM, Henrik Stoerner wrote:
is available at http://www.hswn.dk/beta/post-RC3.patch

[...] 
To install the patch, do:

   cd hobbit-4.0-RC3
   patch -p0 </tmp/post-RC3.patch
   rm Makefile
   ./configure
   make
   su -c "make install"
May be another silly error :-) . This is what I get while trying to
compile

gcc -g -O2 -Wall -Wno-unused -D_REENTRANT -DHAVE_RPCENT -DMAXMSG=32768
-DBBDPORTNUMBER=1984 -I. -I/usr/share/src/hobbit-4.0-RC3/include -o
hobbit-mailack  hobbit-mailack.o ../lib/libbbgen.a -L/usr/local/lib
-lpcre
Undefined                       first referenced
 symbol                             in file
recv                                ../lib/libbbgen.a(sendmsg.o)
getservbyname                       ../lib/libbbgen.a(sendmsg.o)
gethostbyname                       ../lib/libbbgen.a(sendmsg.o)
socket                              ../lib/libbbgen.a(sendmsg.o)
getsockopt                          ../lib/libbbgen.a(sendmsg.o)
connect                             ../lib/libbbgen.a(sendmsg.o)
inet_aton                           ../lib/libbbgen.a(sendmsg.o)
inet_ntoa                           ../lib/libbbgen.a(sendmsg.o)
shutdown                            ../lib/libbbgen.a(sendmsg.o)
ld: fatal: Symbol referencing errors. No output written to
hobbit-mailack
collect2: ld returned 1 exit status
make[1]: *** [hobbit-mailack] Error 1
make[1]: Leaving directory `/usr/share/src/hobbit-4.0-RC3/hobbitd'
make: *** [hobbitd-build] Error 2
I did a 'man getservbyname' and found out I needed to use -lsocket
-lnsl. I added next to -lpcre in the Makefile and my error message
shrinked to this
quoted from Andy France

gcc -g -O2 -Wall -Wno-unused -D_REENTRANT -DHAVE_RPCENT -DMAXMSG=32768
-DBBDPORTNUMBER=1984 -I. -I/usr/share/src/hobbit-4.0-RC3/include -o
hobbit-mailack  hobbit-mailack.o ../lib/libbbgen.a -L/usr/local/lib

-lpcre -lsocket -lnsl
Undefined                       first referenced
 symbol                             in file
inet_aton                           ../lib/libbbgen.a(sendmsg.o)
quoted from Andy France
ld: fatal: Symbol referencing errors. No output written to
hobbit-mailack
collect2: ld returned 1 exit status
make[1]: *** [hobbit-mailack] Error 1
make[1]: Leaving directory `/usr/share/src/hobbit-4.0-RC3/hobbitd'
make: *** [hobbitd-build] Error 2


Now I did a 'man inet' which also recommeds the socket and nsl library
but no go yet.

Let me know if I can go any further. 
quoted from Andy France

Thanks


-- 
Asif Iqbal
PGP Key: 0xE62693C5 KeyServer: pgp.mit.edu
"It is not the strongest of the species that survives, not the most intelligent, but
 the one most responsive to change."    - Charles Darwin
list Asif Iqbal · Wed, 23 Feb 2005 23:55:37 -0500 ·
quoted from Asif Iqbal
On Wed, Feb 23, 2005 at 11:48:51PM, Asif Iqbal wrote:
On Wed, Feb 23, 2005 at 09:24:30PM, Asif Iqbal wrote:
On Thu, Feb 24, 2005 at 12:19:09AM, Henrik Stoerner wrote:
is available at http://www.hswn.dk/beta/post-RC3.patch
[....]
hobbit-mailack
collect2: ld returned 1 exit status
make[1]: *** [hobbit-mailack] Error 1
make[1]: Leaving directory `/usr/share/src/hobbit-4.0-RC3/hobbitd'
make: *** [hobbitd-build] Error 2
Fixed I had to add -lresolv -lsocket -lnsl next to -lpcre in the
quoted from Asif Iqbal
Makefile
Now I did a 'man inet' which also recommeds the socket and nsl library
but no go yet.

Let me know if I can go any further. 

-- 
Asif Iqbal
PGP Key: 0xE62693C5 KeyServer: pgp.mit.edu
"It is not the strongest of the species that survives, not the most intelligent, but
 the one most responsive to change."    - Charles Darwin
list Henrik Størner · Thu, 24 Feb 2005 07:36:08 +0100 ·
quoted from Asif Iqbal
On Wed, Feb 23, 2005 at 09:24:30PM -0500, Asif Iqbal wrote:
On Thu, Feb 24, 2005 at 12:19:09AM, Henrik Stoerner wrote:
is available at http://www.hswn.dk/beta/post-RC3.patch
May be another silly error :-) . This is what I get while trying to
compile
Of course - it needs the net-libraries.

Adding $(NETLIBS) will fix it.


Henrik
list Henrik Størner · Thu, 24 Feb 2005 13:09:05 +0100 ·
quoted from Bruce Lysik
On Wed, Feb 23, 2005 at 03:59:16PM -0800, Bruce Lysik wrote:
And I just got some cpu alerts sent to the onduty address.  
In fact, the 'info' page looks totally wrong for alerts.    
Ack.  Just got another alert I shouldn't have.  (A purple for 
a vmstat monitor which should have only gone to the 'ops' address.)
I've confirmed that this is an issue with the patch.  I rolled back
 to RC3 without patch, and suddenly the alerting definitions that are
 displayed on the 'info' pages make sense again.
Thanks for providing a good test-case with the configuration you use.

The patch you tried was a bit too eager in applying the criteria for
recipients; it used the criteria from the last recipient for all of
the recipients hanging off the same rule (i.e. all lines with
recipients would get the same setup).

I've fixed this, and updated the post-RC3 patch. With the new version,
your configuration results in the correct alert-setup, as far as I can
tell.

Patch is at http://www.hswn.dk/beta/post-RC3.patch (same location as
before). You must un-install the old version of the patch first:

   cd hobbit-4.0-RC3
   patch -R -p0 </tmp/post-RC3.patch.old

then apply the new one:
   patch -p0 </tmp/post-RC3.patch

and then do the "make" "make install".


I'd be very happy if you can confirm that this results in the correct
alert-setup for your hosts. 


Regards,
Henrik
list Henrik Størner · Thu, 24 Feb 2005 13:18:53 +0100 ·
On Thu, Feb 24, 2005 at 03:02:44PM +1300, Andy France wrote:
For my Solaris x86 Hobbit server, I use the Blastwave packages
(http://www.blastwave.org) for installing the libraries required for
compiliation.

Would it be possible to get the following search directories added to the
build scripts to use the Blastwave /opt/csw tree as well?
Done.


Henrik
list Tom Georgoulias · Thu, 24 Feb 2005 08:43:32 -0500 ·
quoted from Henrik Størner
Henrik Stoerner wrote:
Patch is at http://www.hswn.dk/beta/post-RC3.patch (same location as
before). You must un-install the old version of the patch first:

   cd hobbit-4.0-RC3
   patch -R -p0 </tmp/post-RC3.patch.old

then apply the new one:
   patch -p0 </tmp/post-RC3.patch

and then do the "make" "make install".
If I didn't apply the old patch, was I still supposed to delete the Makefile?
quoted from Henrik Størner
I'd be very happy if you can confirm that this results in the correct
alert-setup for your hosts.
I can confirm that my alerts setup in the info page is now displaying properly, which it wasn't when I upgraded to RC3 and before I applied the latest patch.  (having the alert rules show up in info is an awesome feature, BTW.)

Haven't seen any hobbitd_larrd core files either...so far so good.

Tom
list Henrik Størner · Thu, 24 Feb 2005 15:05:36 +0100 ·
quoted from Tom Georgoulias
On Thu, Feb 24, 2005 at 08:43:32AM -0500, Tom Georgoulias wrote:
Henrik Stoerner wrote:
Patch is at http://www.hswn.dk/beta/post-RC3.patch (same location as
before). You must un-install the old version of the patch first:

  cd hobbit-4.0-RC3
  patch -R -p0 </tmp/post-RC3.patch.old

then apply the new one:
  patch -p0 </tmp/post-RC3.patch

and then do the "make" "make install".
If I didn't apply the old patch, was I still supposed to delete the Makefile?
Yes. At least if you're building it on a Linux or *BSD platform.
quoted from Tom Georgoulias

I can confirm that my alerts setup in the info page is now displaying properly, which it wasn't when I upgraded to RC3 and before I applied the latest patch.  (having the alert rules show up in info is an awesome feature, BTW.)

Haven't seen any hobbitd_larrd core files either...so far so good.
Really nice - thanks!


Regards,
Henrik
list Daniel J McDonald · Thu, 24 Feb 2005 08:21:36 -0600 ·
quoted from Henrik Størner
On Thu, 2005-02-24 at 15:05 +0100, Henrik Stoerner wrote:
On Thu, Feb 24, 2005 at 08:43:32AM -0500, Tom Georgoulias wrote:
Henrik Stoerner wrote:
I can confirm that my alerts setup in the info page is now displaying 
properly, which it wasn't when I upgraded to RC3 and before I applied 
the latest patch.  (having the alert rules show up in info is an awesome 
feature, BTW.)

Haven't seen any hobbitd_larrd core files either...so far so good.
Really nice - thanks!
Ditto.  my core files are gone now.

I am still having trouble with the apache graphs.  I finally figured out
that I had to enable extendedstatus in
the /etc/httpd/conf/commonhttpd.conf file, and I'm getting a graph that
finally has data other than NaN on the trends page, but no button on the
main page...  Is that expected?

[mcdonalddj at ldap2 hobbit]$ cat server/etc/bb-hosts
#
# Master configuration file for Hobbit and Big Brother

10.10.9.70   ldap2.austin-energy.net      # BBDISPLAY BBPAGER BBNET bbd http://ldap2.austin-energy.net/ noconn apache=http://localhost/server-status?auto ldap ntp
include bb-hosts-critical
include bb-hosts-network


-- 
Daniel J McDonald, CCIE # 2495, CNX
Austin Energy

user-290ce4e24e19@xymon.invalid
list Henrik Størner · Thu, 24 Feb 2005 15:37:10 +0100 ·
quoted from Daniel J McDonald
On Thu, Feb 24, 2005 at 08:21:36AM -0600, Daniel J McDonald wrote:
I am still having trouble with the apache graphs.  I finally figured out
that I had to enable extendedstatus in
the /etc/httpd/conf/commonhttpd.conf file, and I'm getting a graph that
finally has data other than NaN on the trends page, but no button on the
main page...  Is that expected?
It is, yes. The "apache" test only collects some performance
statistics - much like the "vmstat" or "netstat" reports; it has no
way of going red or yellow. So the performance data is sent using the
BB/Hobbit "data" protocol which doesn't result in a button on the
status webpage.


Henrik