I am writing a new ext called "folders" that will provide a check to see that folders do not exceed a certain size. For example, it will check the folders that contain each user's personal network drive, list them all with their sizes, and indicate any that are over the limit. I have this much worked out and think that I have the red, green, yellow parts of the script working properly.
However, it would be even better if I could store the sizes of these folders within Hobbit/Xymon as data in RRD and have them graphed.
How do I send the data to Hobbit in a format that it will understand as being RRD data and that will be tied to the "folders" test for this host?
Thanks in advance!
Harold Ballinger IT Manager Heritage Healthcare, Inc. (888) 335-2620 | helpdesk (864) 224-3626 | office (864) 224-3093 | fax
[cid:image001.jpg at 01CA2A40.D4045E10]
Visit our website: www.heritage-healthcare.com<http://www.heritage-healthcare.com/>
On Monday 31 August 2009, Harold J. Ballinger wrote:
I am writing a new ext called "folders" that will provide a check to see that folders do not exceed a certain size. For example, it will check the folders that contain each user's personal network drive, list them all with their sizes, and indicate any that are over the limit. I have this much worked out and think that I have the red, green, yellow parts of the script working properly.
However, it would be even better if I could store the sizes of these folders within Hobbit/Xymon as data in RRD and have them graphed.
How do I send the data to Hobbit in a format that it will understand as being RRD data and that will be tied to the "folders" test for this host? If you have the latest xymon version, you can send an extra trends message.
An example:
bb <xymon server> "data <server name>.trends [test.rrd] DS:test:GAUGE:600:0:U 10"
This will create a test.rrd file in the rrd directory on the hobbit server with the numer 10 in the data source named test. In your case, I think it's best to name your rrd's like the rrds used to store the filesystem information.
And of course you need some extra stuff in hobbitgraph.cfg. If you send the folder message, you can include the needed links to the graphs so the graphs are displayed on the folter check page.
Stef
I am writing a new ext called “folders” that will provide a check to see that folders do not exceed a certain size. For example, it will check the folders that contain each user’s personal network drive, list them all with their sizes, and indicate any that are over the limit. I have this much worked out and think that I have the red, green, yellow parts of the script working properly.
However, it would be even better if I could store the sizes of these folders within Hobbit/Xymon as data in RRD and have them graphed.
How do I send the data to Hobbit in a format that it will understand as being RRD data and that will be tied to the “folders” test for this host?
This sounds like a variable number of measurements (one per folder) with a fixed number of parameters per measurement (one (size) or two (size,max)). In that case you should use the '--extra-test' option (see http://www.hswn.dk/hobbit/help/manpages/man8/hobbitd_rrd.8.html). You probably also need a script to build the header of each graph. If you like, I can send you sample scripts.
Within the body of the status message, you'll include the measured values in NCV format, which might be embedded in HTML comment. Note that you might need to change other lines in the status message in order to avoid them to be interpreted as NCV. One way of doing so is to replace each ":" by ":" and each "=" by "=". The script named in "--extra-test" writes those values to RRDs. O, include a line in the status message, before the measurements, stating the number of measurements, that is include a line with "<!-- linecount=N -->" with a suitable value for N.
Kind regards, Wim Nelis.
The NLR disclaimer (http://www.nlr.nl/emaildisclaimer) is valid for NLR e-mail messages.
This is an example of the output that I am working from to build my script: -------------------Output Pasted Below------------------------------ Searching for directories that have exceeded 3000 megabytes Size (mb) Files Directory 2265.08 7567 SUB-TOTAL: E:\USERS\aeads 164.12 6264 SUB-TOTAL: E:\USERS\atimberlake 26.75 2236 SUB-TOTAL: E:\USERS\bdunlap 1092.05 3783 SUB-TOTAL: E:\USERS\ccampbell 989.58 3143 SUB-TOTAL: E:\USERS\ccampbell.old 982.81 3124 SUB-TOTAL: E:\USERS\ccampbell.oldest 89.50 2660 SUB-TOTAL: E:\USERS\ccannady 0.00 0 SUB-TOTAL: E:\USERS\cdavis ! 3459.78 10628 SUB-TOTAL: E:\USERS\dbailey 0.52 6 SUB-TOTAL: E:\USERS\dbarker 1193.83 840 SUB-TOTAL: E:\USERS\ddavis 230.26 2887 SUB-TOTAL: E:\USERS\dlowry 2312.57 3822 SUB-TOTAL: E:\USERS\dpalacio 4.91 230 SUB-TOTAL: E:\USERS\epaschalis 0.00 0 SUB-TOTAL: E:\USERS\ginnmon 1479.80 3302 SUB-TOTAL: E:\USERS\hballinger 9.92 642 SUB-TOTAL: E:\USERS\jcollier 37.52 1155 SUB-TOTAL: E:\USERS\jhaddad 0.00 0 SUB-TOTAL: E:\USERS\jscott 1269.20 3697 SUB-TOTAL: E:\USERS\jtrew 0.00 0 SUB-TOTAL: E:\USERS\kbowser 0.00 0 SUB-TOTAL: E:\USERS\kbradley 552.58 987 SUB-TOTAL: E:\USERS\kdarr 107.06 3177 SUB-TOTAL: E:\USERS\kmast 112.47 1636 SUB-TOTAL: E:\USERS\kmerritt 3.30 164 SUB-TOTAL: E:\USERS\leads ! 3580.77 11199 SUB-TOTAL: E:\USERS\lfrick 537.32 1762 SUB-TOTAL: E:\USERS\lfrick.old ! 9255.93 16732 SUB-TOTAL: E:\USERS\lfrick.oldest 1207.79 6822 SUB-TOTAL: E:\USERS\lhayes 417.43 3142 SUB-TOTAL: E:\USERS\lheiser 83.59 1328 SUB-TOTAL: E:\USERS\lzipsir 2101.66 3466 SUB-TOTAL: E:\USERS\mzatcoff 256.97 2249 SUB-TOTAL: E:\USERS\ncrawford ! 3163.03 5488 SUB-TOTAL: E:\USERS\pshaw 46.40 628 SUB-TOTAL: E:\USERS\seads 0.00 0 SUB-TOTAL: E:\USERS\shawes 905.58 4948 SUB-TOTAL: E:\USERS\swoodard 2.48 189 SUB-TOTAL: E:\USERS\test1 0.03 33 SUB-TOTAL: E:\USERS\_ICONS 37942.57 119936 TOTAL -------------------End Copy/Paste-------------------------- If one of the directories is over the limit that has been defined, the line is indicated with a "!" at the first character position. I will prepend the output with "&green %date% %time%" or "&red %date% %time%" depending on whether or not any "!" are found in the file to indicate the status of the test. I will then replace the "!" with "&red" before storing the file into %BBHOME%\logs\folders. I am considering having the "SUB-TOTAL: " stripped from each line and having the header and footer/total lines stripped as well before sending. This just covers how I will send the test to Xymon.
It sounds like you have some sample scripts that could help me in taking it to the next level and by getting this reporting as data into RRD for graphing. I am very interested in seeing these scripts so that I can work on getting this working 100%.
Thanks so much.
-----Original Message----- From: W.J.M. Nelis [mailto:nelis at nlr.nl] Sent: Tuesday, September 01, 2009 6:32 AM To: hobbit at hswn.dk Subject: Re: [hobbit] Folders RRD info
I am writing a new ext called "folders" that will provide a check to see that folders do not exceed a certain size. For example, it will check the folders that contain each user's personal network drive, list them all with their sizes, and indicate any that are over the limit. I have this much worked out and think that I have the red, green, yellow parts of the script working properly.
However, it would be even better if I could store the sizes of these folders within Hobbit/Xymon as data in RRD and have them graphed.
How do I send the data to Hobbit in a format that it will understand as being RRD data and that will be tied to the "folders" test for this host?
This sounds like a variable number of measurements (one per folder) with a fixed number of parameters per measurement (one (size) or two (size,max)). In that case you should use the '--extra-test' option (see http://www.hswn.dk/hobbit/help/manpages/man8/hobbitd_rrd.8.html). You probably also need a script to build the header of each graph. If you like, I can send you sample scripts.
Within the body of the status message, you'll include the measured values in NCV format, which might be embedded in HTML comment. Note that you might need to change other lines in the status message in order to avoid them to be interpreted as NCV. One way of doing so is to replace each ":" by ":" and each "=" by "=". The script named in "--extra-test" writes those values to RRDs. O, include a line in the status message, before the measurements, stating the number of measurements, that is include a line with "<!-- linecount=N -->" with a suitable value for N.
Kind regards, Wim Nelis.
The NLR disclaimer (http://www.nlr.nl/emaildisclaimer) is valid for NLR e-mail messages.
To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk
You will probably want to approach this the same way that Xymon handles disk sizes, with each disk getting stored in a separate RRD, and with a naming convention that is compatible with hobbitgraph. It will take some fiddling.
There are two ways to "catch" this data on the Hobbit server and both are documented. One is NCV, and the other is to write your own "data parser" as an external script. I prefer the external data parser because it is more flexible than NCV, but NCV is "compiled in" so therefore "supported".
The thing to look out for is this: the number of directories that you want to track may be more than Hobbitgraph can handle. Depending on where you are going with this you might need to look into using other RRD tools for graphing so you can avoid the limits in Hobbitgraph (although it is pretty good).
Are you going to update this every 5 minutes? If not, you will have some other challenges to overcome if you try to leverage the Hobbit data collection/graphing system.
GLH
On 9/1/09, Harold J. Ballinger <hballinger at heritage-healthcare.com> wrote:
This is an example of the output that I am working from to build my script: -------------------Output Pasted Below------------------------------ Searching for directories that have exceeded 3000 megabytes Size (mb) Files Directory 2265.08 7567 SUB-TOTAL: E:\USERS\aeads 164.12 6264 SUB-TOTAL: E:\USERS\atimberlake 26.75 2236 SUB-TOTAL: E:\USERS\bdunlap 1092.05 3783 SUB-TOTAL: E:\USERS\ccampbell 989.58 3143 SUB-TOTAL: E:\USERS\ccampbell.old 982.81 3124 SUB-TOTAL: E:\USERS\ccampbell.oldest 89.50 2660 SUB-TOTAL: E:\USERS\ccannady 0.00 0 SUB-TOTAL: E:\USERS\cdavis ! 3459.78 10628 SUB-TOTAL: E:\USERS\dbailey 0.52 6 SUB-TOTAL: E:\USERS\dbarker 1193.83 840 SUB-TOTAL: E:\USERS\ddavis 230.26 2887 SUB-TOTAL: E:\USERS\dlowry 2312.57 3822 SUB-TOTAL: E:\USERS\dpalacio 4.91 230 SUB-TOTAL: E:\USERS\epaschalis 0.00 0 SUB-TOTAL: E:\USERS\ginnmon 1479.80 3302 SUB-TOTAL: E:\USERS\hballinger 9.92 642 SUB-TOTAL: E:\USERS\jcollier 37.52 1155 SUB-TOTAL: E:\USERS\jhaddad 0.00 0 SUB-TOTAL: E:\USERS\jscott 1269.20 3697 SUB-TOTAL: E:\USERS\jtrew 0.00 0 SUB-TOTAL: E:\USERS\kbowser 0.00 0 SUB-TOTAL: E:\USERS\kbradley 552.58 987 SUB-TOTAL: E:\USERS\kdarr 107.06 3177 SUB-TOTAL: E:\USERS\kmast 112.47 1636 SUB-TOTAL: E:\USERS\kmerritt 3.30 164 SUB-TOTAL: E:\USERS\leads ! 3580.77 11199 SUB-TOTAL: E:\USERS\lfrick 537.32 1762 SUB-TOTAL: E:\USERS\lfrick.old ! 9255.93 16732 SUB-TOTAL: E:\USERS\lfrick.oldest 1207.79 6822 SUB-TOTAL: E:\USERS\lhayes 417.43 3142 SUB-TOTAL: E:\USERS\lheiser 83.59 1328 SUB-TOTAL: E:\USERS\lzipsir 2101.66 3466 SUB-TOTAL: E:\USERS\mzatcoff 256.97 2249 SUB-TOTAL: E:\USERS\ncrawford ! 3163.03 5488 SUB-TOTAL: E:\USERS\pshaw 46.40 628 SUB-TOTAL: E:\USERS\seads 0.00 0 SUB-TOTAL: E:\USERS\shawes 905.58 4948 SUB-TOTAL: E:\USERS\swoodard 2.48 189 SUB-TOTAL: E:\USERS\test1 0.03 33 SUB-TOTAL: E:\USERS\_ICONS 37942.57 119936 TOTAL -------------------End Copy/Paste-------------------------- If one of the directories is over the limit that has been defined, the line is indicated with a "!" at the first character position. I will prepend the output with "&green %date% %time%" or "&red %date% %time%" depending on whether or not any "!" are found in the file to indicate the status of the test. I will then replace the "!" with "&red" before storing the file into %BBHOME%\logs\folders. I am considering having the "SUB-TOTAL: " stripped from each line and having the header and footer/total lines stripped as well before sending. This just covers how I will send the test to Xymon.
It sounds like you have some sample scripts that could help me in taking it to the next level and by getting this reporting as data into RRD for graphing. I am very interested in seeing these scripts so that I can work on getting this working 100%.
Thanks so much.
-----Original Message----- From: W.J.M. Nelis [mailto:nelis at nlr.nl] Sent: Tuesday, September 01, 2009 6:32 AM To: hobbit at hswn.dk Subject: Re: [hobbit] Folders RRD info
I am writing a new ext called "folders" that will provide a check to see that folders do not exceed a certain size. For example, it will check the folders that contain each user's personal network drive, list them all with their sizes, and indicate any that are over the limit. I have this much worked out and think that I have the red, green, yellow parts of the script working properly.
However, it would be even better if I could store the sizes of these folders within Hobbit/Xymon as data in RRD and have them graphed.
How do I send the data to Hobbit in a format that it will understand as being RRD data and that will be tied to the "folders" test for this host?
This sounds like a variable number of measurements (one per folder) with a fixed number of parameters per measurement (one (size) or two (size,max)). In that case you should use the '--extra-test' option (see http://www.hswn.dk/hobbit/help/manpages/man8/hobbitd_rrd.8.html). You probably also need a script to build the header of each graph. If you like, I can send you sample scripts.
Within the body of the status message, you'll include the measured values in NCV format, which might be embedded in HTML comment. Note that you might need to change other lines in the status message in order to avoid them to be interpreted as NCV. One way of doing so is to replace each ":" by ":" and each "=" by "=". The script named in "--extra-test" writes those values to RRDs. O, include a line in the status message, before the measurements, stating the number of measurements, that is include a line with "<!-- linecount=N -->" with a suitable value for N.
Kind regards, Wim Nelis.
The NLR disclaimer (http://www.nlr.nl/emaildisclaimer) is valid for NLR e-mail messages.
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
-- Disclaimer: 1) all opinions are my own, 2) I may be completely wrong, 3) my advice is worth at least as much as what you are paying for it, or your money cheerfully refunded.
On Tuesday 01 September 2009, Greg Hubbard wrote:
You will probably want to approach this the same way that Xymon handles disk sizes, with each disk getting stored in a separate RRD, and with a naming convention that is compatible with hobbitgraph. It will take some fiddling.
There are two ways to "catch" this data on the Hobbit server and both are documented. One is NCV, and the other is to write your own "data parser" as an external script. I prefer the external data parser because it is more flexible than NCV, but NCV is "compiled in" so therefore "supported". There is 3: you can send a trends message, see my previous answer. This is much more flexible then NVC and you don't have to write an external parser. So the script send the normal message and an other trends message.
Stef
participants (4)
-
glh.forums＠gmail.com
-
hballinger＠heritage-healthcare.com
-
nelis＠nlr.nl
-
stef.coene＠docum.org