March 2008 - Xymon - Xymon mailman web

Securing CGI secure by monitored hosts
by bewhite＠fellowes.com 14 Mar '08

14 Mar '08

Hi all, All these questions about securing hobbit has made me think about our hobbit set-up. Is there a way to secure the actions taken by scripts in the secure CGI directory to acting on specific listings in the bb-hosts file? We have one group responsible for printers which is completely different from the group responsible for servers. We would like to give the group responsible for printers the ability to put printers in maintenance mode, but would not want them to have the same access to servers. We are running a generic hobbit section of the httpd.conf file with users defined via the htpasswd command. Thanks, Bruce ---------------------------------------------------- Note: The information contained in this message may be privileged and confidential and protected from disclosure. If the reader of this message is not the intended recipient, or an employee or agent responsible for delivering this message to the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by replying to the message and deleting it from your computer. Thank you. Fellowes, Inc. ----------------------------------------------------

2 1

Current snapshot - compilation error
by alevin＠reveredata.com 14 Mar '08

14 Mar '08

OS: FreeBSD 7 Error: .... In file included from do_rrd.c:517: rrd/do_memory.c:24: error: static declaration of 'string_compare' follows non-static declaration /w1/src/hobbit/hobbit/snapshot/include/../lib/rbtr.h:70: error: previous declaration of 'string_compare' was here In file included from do_rrd.c:520: rrd/do_netstat.c: In function 'do_netstat_rrd': ..... Can be fixed by removing static int string_compare in hobbitd/rrd/do_memory.c : =================================================================== --- hobbitd/rrd/do_memory.c (revision 7646) +++ hobbitd/rrd/do_memory.c (working copy) @@ -21,10 +21,12 @@ */ RbtHandle memhosts; int memhosts_init = 0; -static int string_compare(void *a, void *b) { +/* +static int string_compare(void *a, void *b) +{ return strcmp((char *)a, (char *)b); } - +*/ static int get_mem_percent(char *l) { char *p; ====================================================================== __________________________________ Alexandre Levin

3 2

Snapshot differs
by lars.ebeling＠leopg9.no-ip.org 14 Mar '08

14 Mar '08

Shouldn't the content of hobbit-snapshot.tar.gz be the same as the content of the snapshot directory? At least version.h differs. -- Regards Lars Ebeling http://leopg9.no-ip.org Hobbithobbyist "It is better to keep your mouth shut and appear stupid than to open it and remove all doubt." -- Mark Twain

2 1

ncv graphing issue with custom script.
by jeff.vander＠gmail.com 14 Mar '08

14 Mar '08

Hello all, wondering if anyone can help with a graphing issue I am having... i have a requirement to monitor some unc shares from some windows boxes. i created a script to map each share, get the % free and report this back to hobbit. script works fine but the graphing is giving me problems. the script will send this to hobbit: green 3/30/2030 2:54:19 PM Percent full status of shares that were checked: \\<someserver1>\<some_share3> : 23 \\<someserver2>\<some_share4> : 0 etc : etc etc : etc i guess my question is this type of data suitable for ncv graphing considering that the amount of shares to check can be 10 or 20. my problem is that the "name" portion of the ncv is always changing... i did get graphing to work but the service section in the hobbitgraph.cfg had to have each share entered individually...

3 7

Securing Hobbit from visitors
by josh＠imaginenetworksllc.com 13 Mar '08

13 Mar '08

I am curious to see how the crew here on the mailing list secures their Hobbit from the outside world. I need to have the WWW pages visible from every IP but only from certain people, therefor I need to use users and passwords. Our Hobbitmon is viewed via cell phones and computers (IE and Firefox) and protected by an HTTP(S) login currently. The problem is that with three different Directory statements in httpd.conf, you need to login three times every time you restart Firefox. Also, how many businesses have Hobbitmon wide open for the viewing, such as Henrik's demo, if any? -- Josh Luthman Office: 937-552-2340 Direct: 937-552-2343 1100 Wayne St Suite 1337 Troy, OH 45373 Those who don't understand UNIX are condemned to reinvent it, poorly. --- Henry Spencer

8 22

Possible to have Hobbit retain and use cookies?
by martin.e.rogers＠googlemail.com 13 Mar '08

13 Mar '08

We have an issue in using Hobbit to monitor a system of WebLogic and Oracle application servers. Hobbit is monitoring each of our 4 WebLogic servers on a regular (30/60 second) basis. By default, WebLogic creates a unique HTTPSession when each request is served; this is keyed with a JSESSIONID cookie which is sent back with the reply. A browser, or our load generator (Grinder) retains the cookie. Hobbit seems to be discarding these cookies; the result is that each request is generating a new Session. This has little performance overhead, but serves to confuse our WebLogic statistics; we can not tell how many 'real' users there are logged on to our system, because they are masked by 700+ 'hobbit' sessions. It is possible to have Hobbit retain these cookies and send with subsequent requests to the same URL? Martin

5 7

[hobbit] alternate page graphs custom menu
by jokketan＠hotmail.com 13 Mar '08

13 Mar '08

I had already tried the same thing but i don't succeed in getting hobbit to show the alternate templates for the current graphs. Even when i add the --templates=/wherever/you/want directive to this line in hobbitcgi.cnf ( i know this would change the main site also, it's just a test )# hobbitsvc.cgi options for a normal (current) status displayCGI_SVC_OPTS="--env=/usr/lib/hobbit/server/etc/hobbitserver.cfg --no-svcid --history=top"I also tried adding another variable ($ALT_OPTS) that i created myself in the .sh script that generates the page and then adding it to the bb-hostsvc.sh like this: . /usr/local/hobbit/server/etc/hobbitcgi.cfg exec /usr/local/hobbit/server/bin/hobbitsvc.cgi $CGI_SVC_OPTS $ALT_OPTSWhich basically also does nothing...Can anyone shed some light on this??? Seems like i am missing something obvious...Thanks,Jonathan Van Liefde References: alternate page graphs custom menu From: Jonathan van liefde Prev by Date: Testing Hobbit Client Log "IGNORE" filters Next by Date: RE: [hobbit] Question Previous by thread: alternate page graphs custom menu Next by thread: Question about hobbit alert acknowledgement, delay of next alert time Index(es): DateThread To: <hobbit (at) hswn.dk>Subject: RE: [hobbit] alternate page graphs custom menuFrom: "Tom Moore" <Tom.Moore (at) sas.com>Date: Mon, 30 Jul 2007 16:42:04 -0400References: <BAY139-W17B66F209315AF69B9FFAAD4EE0 (at) phx.gbl>Thread-index: AcfS6INBn1czzAgURTOpkFRYPh5mBAAAJ72QThread-topic: [hobbit] alternate page graphs custom menu That's exactly the problem I am having and posted to the list a few weeks ago. It seems that the actual cgi scripts that run when you click on an item use the default header sets, rather than the ones specified by your --templates or --pageset settings with bbgen. The first page on the pageset does use the custom_header and custom_footer files, but if you click on an item like "cpu", it runs a cgi script which generates a new page using the default templates instead of your custom ones. For example, most of the items call "bb-hostsvc.sh". which is just a wrapper that takes arguments and runs . /usr/local/hobbit/server/etc/hobbitcgi.cfg exec /usr/local/hobbit/server/bin/hobbitsvc.cgi $CGI_SVC_OPTS Now, optimally you should be able to put a custom $CGI_SVC_OPTS definition in whatever script generates your alternate pages, such that CGI_SVC_OPTS=--templates=/whatever/directory/with/your/new/templatefiles Unfortunately, even this doesn't seem to work. The key is to get your custom sub-pages to use custom templates for the graph pages and such that are dynamically generated by the CGI files. From: Jonathan van liefde [mailto:jokketan (at) hotmail.com] Sent: Monday, July 30, 2007 1:40 PMTo: hobbit (at) hswn.dkSubject: [hobbit] alternate page graphs custom menu Hi,My problem is the following:We are using the monitoring tool for a helpdesk, every member of the helpdesk has full access to every hobbit page.Local guys should have access to a limited area of the hobbit page.So i have build an alternate page set with all different sections and restricted access with a .htaccess file. This alternate page set uses a different set of headers and footers (without the menu) to prevent local guys from wandering off.But when they look at for example the conn graph for a given host, hobbit uses the bb_header and bb_footer which still has the menu in it.Is there any way to tell hobbit to use the alternate page set header and footer for the graphs of a host when they are accessed through the alternate page set?Thanks for your help,Jonathan Van Liefde.________________________________het ultieme online mailprogramma! Ontdek Windows Live Hotmail, <http://get.live.com/mail/overview> References: alternate page graphs custom menu From: Jonathan van liefde Prev by Date: Testing Hobbit Client Log "IGNORE" filters Next by Date: RE: [hobbit] Question Previous by thread: alternate page graphs custom menu Next by thread: Question about hobbit alert acknowledgement, delay of next alert time Index(es): DateThread _________________________________________________________________ Ontdek Windows Live Hotmail, het ultieme online mailprogramma! http://get.live.com/mail/overview

2 1

Using Hobbit to monitor a specific port for availability?
by CallahanT＠tessco.com 13 Mar '08

13 Mar '08

How would I get hobbit to monitor a specific port to ensure it is listening? For instance, I want to monitor a Java program that listens on port 8888, but just want hobbit to check if it can connect, and then disconnect. How could I do this? Thanks, Tom Callahan

5 7

Getting red for disk /cdrom being 100% full.
by rmcgraw＠purdue.edu 12 Mar '08

12 Mar '08

INFO: SunOS zorn 5.10 Generic_120011-14 sun4u sparc SUNW,Sun-Fire-280R Running hobbit 4.2.0. The host zorn it the hobbit server. Below if information being returned under the "disk" heading. The problem is that hobbit is reporting "red" because RED /cdrom/sol_10_807_sparc/s1 (100% used) has reached the PANIC level (95%) Filesystem kbytes used avail capacity Mounted on /dev/md/dsk/d0 16526762 12802540 3558955 79% / /dev/md/dsk/d3 8263373 6865699 1315041 84% /var /dev/dsk/c3t102d0s6 906514080 725338492 172110448 81% /export/users /dev/dsk/c3t102d1s6 98989396 75838865 22160638 78% /export/pkgs /dev/dsk/c3t102d1s7 53621849 3167622 49918009 6% /export/csw /dev/dsk/c3t102d2s6 98989396 31142015 66857488 32% /export/jumpstart /dev/dsk/c3t102d2s7 53621849 52716 53032915 1% /export/patchsvr /dev/dsk/c3t102d3s6 50869921 20223963 30137259 41% /export/varmail /dev/dsk/c3t102d4s6 832492688 326229172 497938590 40% /export/loc1 /vol/dev/dsk/c0t6d0/sol_10_807_sparc/s6 31 2 26 8% /cdrom/sol_10_807_sparc/s6 /vol/dev/dsk/c0t6d0/sol_10_807_sparc/s5 31 2 26 8% /cdrom/sol_10_807_sparc/s5 /vol/dev/dsk/c0t6d0/sol_10_807_sparc/s4 31 2 26 8% /cdrom/sol_10_807_sparc/s4 /vol/dev/dsk/c0t6d0/sol_10_807_sparc/s3 31 2 26 8% /cdrom/sol_10_807_sparc/s3 /vol/dev/dsk/c0t6d0/sol_10_807_sparc/s2 31 2 26 8% /cdrom/sol_10_807_sparc/s2 /vol/dev/dsk/c0t6d0/sol_10_807_sparc/s1 174639 157917 0 100% /cdrom/sol_10_807_sparc/s1 swap 14135032 1408 14133624 1% /etc/svc/volatile swap 14137184 3560 14133624 1% /tmp swap 14133664 40 14133624 1% /var/run I have modified the df command client/bin/hobbitclient-sunos.sh to strip this out and it works correctly for the trends. How/where do I modify hobbit to skip the /vol so it does not get reported as being 100% full. Thanks Robert -------------------------------------------------------------------- Robert P. McGraw, Jr. Manager, Computer System EMAIL: rmcgraw at purdue.edu Purdue University ROOM: MATH-807 Department of Mathematics PHONE: (765) 494-6055 150 N. University Street West Lafayette, IN 47907-2067

4 4

Lines to flag red or yellow in msgs
by josh＠imaginenetworksllc.com 12 Mar '08

12 Mar '08

Is it possible for me to define lines that appear in msgs to activate a red or yellow alert? I have a bond0 interface (bonded interface consisting of eth0 and eth1) and if I unplug and plug the cat5 cable (simulating it being down) I get these messages: Mar 10 14:48:15 hostname kernel: e100: eth1: e100_watchdog: link down Mar 10 14:48:15 hostname kernel: bonding: bond0: link status definitely down for interface eth1, disabling it Mar 10 14:48:17 hostname kernel: e100: eth1: e100_watchdog: link up, 100Mbps, full-duplex Mar 10 14:48:17 hostname kernel: bonding: bond0: link status definitely up for interface eth1. The /var/log/messages is definitely making it to the Hobbit Server as I can see the text on the msgs web page, but it isn't flagging them as a bad message. Josh -- Josh Luthman Office: 937-552-2340 Direct: 937-552-2343 1100 Wayne St Suite 1337 Troy, OH 45373 Those who don't understand UNIX are condemned to reinvent it, poorly. --- Henry Spencer

3 7