Hello,
Is anyone aware of a way to generate reports via the CLI that show, for example, a host's CPU average over the last 30 days?
I can get the data in graph form via the xymon web portal's "Metrics Reports" page, but need to gather this data point for a large number of hosts.
I also looked at the manpage for 'xymon', and xymondboard appears to gather current stats, but I didn't see a way to gather a date range.
Thank you,
Bill Howe howe.bill at gmail.com <http://www.linkedin.com/in/whowe>
Bill Howe writes:
Is anyone aware of a way to generate reports via the CLI that show, for example, a host's CPU average over the last 30 days?
I can get the data in graph form via the xymon web portal's "Metrics Reports" page, but need to gather this data point for a large number of hosts.
I also looked at the manpage for 'xymon', and xymondboard appears to gather current stats, but I didn't see a way to gather a date range.
It depends upon what you mean by "a host's CPU average over the last 30 days". Do you mean one number that's the 30 day average? 30 numbers that each show a 1 day average? Every 5 minute number for 30 days?
In any case, the likely answer is to use rrdfetch and pull the data directly from the rrd files. Do read the RESOLUTION INTERVAL section of the rrdfetch manpage. If you don't have your boundries correct you won't get the answers you're expecting. Note that you likely don't have 5 minute resolution going back anywhere near 30 days. (By default you'll likely have 2 hour resolution for that period.)
For example, to get a 24hr average for each of the last 30 days: TIME=$(date +%s) RRDRES=86400 rrdtool fetch la.rrd AVERAGE -r $RRDRES -e $(($TIME/$RRDRES*$RRDRES)) -s e-30d
Also note that you need to adjust the results. For example:
7.3520000000e+01 = .73520000000 1.0855666667e+02 = 1.0855666667
There is no way to get this directly afaik. But you can extract those values from your rrd files with the start and end arg with rrdtool.
For example to get the average load for a server for the last 15 days just do:
rrdtool fetch la.rrd AVERAGE -s now-15d -e now | gawk '!/nan/{sum+=$2};END {printf("Average load: %.1f\n",sum/NR)}'
For a bunch of servers loop like this
for server in $list do rdtool fetch $XYMONRRDS/$server/la.rrd AVERAGE -s now-15d -e now|gawk '!/nan/{sum+=$2};END{printf("Average load server %s: %.1f \n","'"$server"'",sum/NR)}' done
Norbert
From: Bill Howe <howe.bill at gmail.com> To: Xymon Mailing List <xymon at xymon.com> Date: 03/01/2019 04:55 PM Subject: [Xymon] CPU Average for 1 month via CLI Sent by: "Xymon" <xymon-bounces at xymon.com>
Hello,
Is anyone aware of a way to generate reports via the CLI that show, for example, a host's CPU average over the last 30 days?
I can get the data in graph form via the xymon web portal's "Metrics Reports" page, but need to gather this data point for a large number of hosts.
I also looked at the manpage for 'xymon', and xymondboard appears to gather current stats, but I didn't see a way to gather a date range.
Thank you,
Bill Howe howe.bill at gmail.com_______________________________________________ Xymon mailing list Xymon at xymon.com https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.xymon.com_mailman_...
Thank you Michael and Norbert. I will look into rrdtool fetch!
Regards,
Bill Howe howe.bill at gmail.com <http://www.linkedin.com/in/whowe>
On Fri, Mar 1, 2019 at 12:04 PM Norbert Kriegenburg < norbert.kriegenburg at de.ibm.com> wrote:
There is no way to get this directly afaik. But you can extract those values from your rrd files with the start and end arg with rrdtool.
For example to get the average load for a server for the last 15 days just do:
rrdtool fetch la.rrd AVERAGE -s now-15d -e now | gawk '!/nan/{sum+=$2};END{printf("Average load: %.1f\n",sum/NR)}'
For a bunch of servers loop like this
for server in $list do rdtool fetch $XYMONRRDS/$server/la.rrd AVERAGE -s now-15d -e now|gawk '!/nan/{sum+=$2};END{printf("Average load server %s: %.1f\n","'"$server"'",sum/NR)}' done
Norbert
[image: Inactive hide details for Bill Howe ---03/01/2019 04:55:12 PM---Hello, Is anyone aware of a way to generate reports via the CLI]Bill Howe ---03/01/2019 04:55:12 PM---Hello, Is anyone aware of a way to generate reports via the CLI that show, for
From: Bill Howe <howe.bill at gmail.com> To: Xymon Mailing List <xymon at xymon.com> Date: 03/01/2019 04:55 PM Subject: [Xymon] CPU Average for 1 month via CLI Sent by: "Xymon" <xymon-bounces at xymon.com>
Hello,
Is anyone aware of a way to generate reports via the CLI that show, for example, a host's CPU average over the last 30 days?
I can get the data in graph form via the xymon web portal's "Metrics Reports" page, but need to gather this data point for a large number of hosts.
I also looked at the manpage for 'xymon', and xymondboard appears to gather current stats, but I didn't see a way to gather a date range.
Thank you,
Bill Howe *howe.bill at gmail.com* <howe.bill at gmail.com> <http://www.linkedin.com/in/whowe>
Xymon mailing list Xymon at xymon.com http://lists.xymon.com/mailman/listinfo/xymon
participants (3)
-
howe.bill＠gmail.com
-
mtpins＠nndev.org
-
norbert.kriegenburg＠de.ibm.com