September 2005 - Xymon - Xymon mailman web

10-11pm RRD gap in graphs (4.1.1)
by craig＠cookitservices.com 20 Sep '05

20 Sep '05

I have an odd issue... >From around 10pm at night to around 11pm (EST) I have a gap in my CPU and Processes/Users graphs. I haven't seen any reasons in the hobbit logs. Other RRD graphs are fine (disk, vmstat, etc) This is not just one night, but every night. I am using the hobbit client on the hobbit server, and hobbit client on another linux box (both are CentOS 4). Has anyone else seen this? What is significant about CPU, Proc/Users graphs? Thanks Craig Cook -- Systems Monitoring Consulting and Support Services http://www.cookitservices.com

2 1

[hobbit] problem with disk test display
by henrik＠hswn.dk 19 Sep '05

19 Sep '05

On Mon, Sep 19, 2005 at 04:50:54PM -0400, Larry.Barber at usda.gov wrote: > Henrik, it appears that having partitions removed from a Linux/Unix > system messes up the disk drive display. The disk test continues to > display the removed partitions (at least in the legend), but at the > expense of not displaying all of the current partitions. In order to > display all of the current partitions you have to manually remove > the .rrd files that correspond to the removed partitions. Ah yes - this is probably the same bug that Pat Vaughan reported earlier today. > Evidently the program just goes through the .rrd files until it has > read the same number of files as there are lines in the report from > the client. That's precisely what it does. > It might be better to read all the .rrd files, regardless > of number, or to check that the .rrd file corresponds with a line in > the disk test data. That is a bit more work, and for sites running hobbitd_filestore with the "--html" option to save the HTML status log every time a status message is received it will be a performance killer. But I don't know of anyone who does that - it really doesn't make sense with Hobbit. Next problem is that the hobbitsvc CGI that generates the HTML for the status logs doesn't know anything about how RRD files map to graphs. That's done by the hobbitgraph CGI.... So it's a simple problem that requires some re-shuffling to fix. Henrik

1 0

problem with disk test display
by Larry.Barber＠usda.gov 19 Sep '05

19 Sep '05

Henrik, it appears that having partitions removed from a Linux/Unix system messes up the disk drive display. The disk test continues to display the removed partitions (at least in the legend), but at the expense of not displaying all of the current partitions. In order to display all of the current partitions you have to manually remove the .rrd files that correspond to the removed partitions. Evidently the program just goes through the .rrd files until it has read the same number of files as there are lines in the report from the client. It might be better to read all the .rrd files, regardless of number, or to check that the .rrd file corresponds with a line in the disk test data. Thanks, Larry Barber

1 0

Feature request
by bigdan＠gmail.com 19 Sep '05

19 Sep '05

On my old bb server, I could click on any server's trends and it would take to a page with the history displayed on the graphs. More importantly, up top it would give me a lot of options like "<this trend> for all servers in <category>" or "<this trend> for all servers". Using this I could view memory usage/ping times/disk usage (etc.) for every server I have all on 1 page. Is there a workaround to get this? Is it possible I can get this in as a feature request? Thanks Dan

2 2

Hobbit, Net-snmp snmptrap,and HPOV
by Mark.Coupe＠state.ma.us 19 Sep '05

19 Sep '05

All, I'm trying to push alarms from Hobbit to HPOV. We have a custom script that is invoking Netsnmp's snmptrap. The snmp traps are reaching the HP Node Manager Event Browser correctly. However, the OV operations trap condition is not recognizing the trap. The trap is being displayed as "unmatched" in the Message Browser even though there is a condition set up for that specific trap. When we look at the unmatched trap, it appears that the OID is not being recognized and an '0' is inserted, for example instead of .1.3.6.1.4.1.7058.0.XX, we are seeing .0.0.XX. Any guidance would be greatly appreaciated. Thanks, -Mark Coupe

4 4

[hobbit] fping multiple interfaces on one host
by toaster＠imber.com 19 Sep '05

19 Sep '05

HTF did I miss that! Thanks! Lee On Sep 19, 2005, at 10:00 AM, Larry.Barber at usda.gov wrote: >> From the bb-hosts man page: >> > > > conn={best,|worst,}IP1[,IP2...] > This adds additional IP-adresses that are pinged > during the nor- > mal "conn" test. So the normal "conn" test must be > enabled (the > default) before this tag has any effect. The IP- > adresses listed > here are pinged in addition to the main IP-address. > > When multiple IP's are pinged, you can choose if ALL > IP's must > respond (the "worst" method), or AT LEAST one IP > must respond > (the "best" setting). All of the IP's are reported > in a single > "conn" status, whose color is determined from > the result of > pinging the IP's and the best/worst setting. The > default method > is "best" - so it will report green if just one > of the IP's > respond to ping. > > > Thanks, > Larry Barber > > > On Mon, 2005-09-19 at 08:50 -0500, toaster at imber.com wrote: > > > Hi, > > We have one last custom script to move over from an old BB > server to > a Hobbit :) > > It was a script to ping all the interfaces of a host and show the > results in a "IP" column. > > With fping you can specify multiple hosts to ping. Is there a > simple > way to have hobbit do this for us already? > Ideally to show the ping results in the "conn" column. > > I have attached a screenshot of what the current script shows. > > Thanks, > > Lee > > > > > > To unsubscribe from the hobbit list, send an e-mail to > hobbit-unsubscribe at hswn.dk > > >

1 0

[hobbit] fping multiple interfaces on one host
by Larry.Barber＠usda.gov 19 Sep '05

19 Sep '05

>From the bb-hosts man page: conn={best,|worst,}IP1[,IP2...] This adds additional IP-adresses that are pinged during the nor- mal "conn" test. So the normal "conn" test must be enabled (the default) before this tag has any effect. The IP-adresses listed here are pinged in addition to the main IP-address. When multiple IP's are pinged, you can choose if ALL IP's must respond (the "worst" method), or AT LEAST one IP must respond (the "best" setting). All of the IP's are reported in a single "conn" status, whose color is determined from the result of pinging the IP's and the best/worst setting. The default method is "best" - so it will report green if just one of the IP's respond to ping. Thanks, Larry Barber On Mon, 2005-09-19 at 08:50 -0500, toaster at imber.com wrote: Hi, We have one last custom script to move over from an old BB server to a Hobbit :) It was a script to ping all the interfaces of a host and show the results in a "IP" column. With fping you can specify multiple hosts to ping. Is there a simple way to have hobbit do this for us already? Ideally to show the ping results in the "conn" column. I have attached a screenshot of what the current script shows. Thanks, Lee

1 0

fping multiple interfaces on one host
by toaster＠imber.com 19 Sep '05

19 Sep '05

Hi, We have one last custom script to move over from an old BB server to a Hobbit :) It was a script to ping all the interfaces of a host and show the results in a "IP" column. With fping you can specify multiple hosts to ping. Is there a simple way to have hobbit do this for us already? Ideally to show the ping results in the "conn" column. I have attached a screenshot of what the current script shows. Thanks, Lee

1 0

Fw: [hobbit] Error compiling on HP-UX
by lars.ebeling＠leopg9.no-ip.org 17 Sep '05

17 Sep '05

Hi, still getting this. But after removing 4 lines in do_vmstat.c the make finish. Regards Lars ----- Original Message ----- From: lars ebeling To: hobbit Sent: Thursday, September 15, 2005 5:34 PM Subject: [hobbit] Error compiling on HP-UX Got this(the same as Robert Watson): cc -g -O -D_REENTRANT -DHPUX -I. -I/home/hobbit/snapshot/include -I/usr/local/in clude -I/usr/local/include -I/usr/local/include -c -o do_rrd.o do_rrd.c In file included from do_rrd.c:224: rrd/do_vmstat.c: In function édo_vmstat_rrd': rrd/do_vmstat.c:275: error: évmstat_linux22_layout' undeclared (first use in thi s function) rrd/do_vmstat.c:275: error: (Each undeclared identifier is reported only once rrd/do_vmstat.c:275: error: for each function it appears in.) rrd/do_vmstat.c:279: error: évmstat_rhel3_layout' undeclared (first use in this function) makeÄ1Å: *** Ädo_rrd.oÅ Error 1 makeÄ1Å: Leaving directory é/home/hobbit/snapshot/hobbitd' make: *** Ähobbitd-buildÅ Error 2 Regards Lars Ebeling http://leopg9.no-ip.org Hobbithobbyist "Never run after your own hat. Others will be delighted to do it; why spoil their fun?" -- Mark Twain

1 0

[hobbit] Very Weird ping issue
by wes.neal＠mci.com 17 Sep '05

17 Sep '05

Does it need to be in " " like that? I added a .default. # testip and it did not work. Wes -----Original Message----- From: Larry.Barber at usda.gov [mailto:Larry.Barber at usda.gov] Sent: Friday, September 16, 2005 5:59 PM To: hobbit at hswn.dk Subject: RE: [hobbit] Very Weird ping issue You will have to ask Henrik about why Hobbit doesn't use the bb-hosts settings. You don't have to enter testip on every line, you can just define a ".default." host at the the top of your bb-hosts file and put testip in there. Thanks, Larry Barber On Fri, 2005-09-16 at 16:44 -0500, wes.neal at mci.com wrote: > Larry you are correct! Thanks! > > My next question is why is it using DNS instead of the bb-hosts file? > Is > there a way to make it only use the bb-hosts file and not rely on DNS > at > all? I always thought it used bb-hosts first. I don't want to have > to add > testip to every line in the file if possible. > > Wes > > -----Original Message----- > From: Larry.Barber at usda.gov [mailto:Larry.Barber at usda.gov] > Sent: Friday, September 16, 2005 5:32 PM > To: hobbit at hswn.dk > Subject: Re: [hobbit] Very Weird ping issue > > What happens if you put "testip" in your bb-hosts file? > > I suspect you might find that the names are being resolved > incorrectly > by DNS, try an nslookup from you BBNET machine, and see what you get. > > Thanks, > Larry Barber > > > On Fri, 2005-09-16 at 16:15 -0500, wes.neal at mci.com wrote: > > I am having a problem with out of the 600 or so hosts I have right > now > > and > > it is very odd. > > > > The four hosts in bb-hosts look like this: > > > > 166.47.52.23 MIREGA-XYP # > > conn > > 166.46.11.1 NRKANG-XYP # > > conn > > 166.42.132.19 ALMMGA-AI > #conn > > 166.39.161.143 ORLLFL-AI # > conn > > > > The all 4 show up as red on the web page, but with not those IP > > addresses. > > As a matter of fact the IPs they show as aren't anywhere in my > > bb-hosts > > file. > > The emails that are generated even have the wrong IPs as well, they > > look > > like this: > > > > red Fri Sep 16 10:09:39 2005 conn NOT ok > > Service conn on ALMMGA-AI is not OK : Host does not respond to > ping > > System unreachable for 9552 poll periods (3195617 seconds) > > &red 10.8.2.126 is unreachable > > > > red Fri Sep 16 10:09:39 2005 conn NOT ok > > Service conn on MIREGA-XYP is not OK : Host does not respond to > ping > > System unreachable for 9552 poll periods (3195617 seconds) > > &red 166.47.118.231 is unreachable > > > > As you can see those aren't the IPs in the bb-hosts file. I have > > commented > > out the lines in the bb-hosts file and done a: > > > > bb 127.0.0.1 "drop MIREGA-XYP" for all of the hosts. This got rid > of > > them, > > but when I put them back in, they came right back red with the same > > bad IPs. > > > > What could possibly be causing this? > > > > Thanks > > Wes > > > > > > > > > > > > To unsubscribe from the hobbit list, send an e-mail to > > hobbit-unsubscribe at hswn.dk > > > > > > > > > > To unsubscribe from the hobbit list, send an e-mail to > hobbit-unsubscribe at hswn.dk > > > > To unsubscribe from the hobbit list, send an e-mail to > hobbit-unsubscribe at hswn.dk > > > To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk

2 1