Configuring an Alert

[Manny Zachodin]

Hi,

I wrote a small TCL program to generate an email every time a thread goes down.  I tested the program from a TCL prompt and it is working fine (generates an email).  To test the Alert, I am bringing down a thread but the Alert doesn

Topic

[Robert Kersemakers]

Hi Manny,

Instead of ‘Duration: once’ try ‘Duration: N seconds’ or ‘Duration: N minutes’ and specify the N value. This should trigger the alert.

Zuyderland Medisch Centrum; Heerlen/Sittard; The Netherlands

[James Cobane]

Manny,

Also take a look at the alert logs (both the .err and the .log) in the $HCISITEDIR/exec/hcimonitord directory.  You can then see if it is firing or erroring.

Hope this helps.

Jim Cobane

Henry Ford Health

[Russ Ross]

I noticed that in your Alert Hosts

Code:

#
#   ***********************************************************************************
#   touch default.alrt files for each site to retrigger alerts each hour***************
#   ***********************************************************************************
#
#     ———————————————
#     Everyday 8:00 AM until 9:00 PM
#     ———————————————
#
#
00 08-21 * * * /bin/ksh -c ‘eval . ~/.profile.cron /quovadx/qdx5.2/integrator siteProto; $HCIROOT/mda_global/scripts/touch_default_alrt.ksh’

Here is the touch_default_alrt.ksh scripts

Code:

#!/usr/bin/ksh

# touch_default_alrt.ksh
# Touch default.alrt files in all key, non-batch prod sites.
# Touching this file re-triggers alerts.

# Presumably this will run from cron at some frequency,
# such as once an hour Monday through Friday, from 7 AM until 9 PM.
clear
# List all sites in prod,
# then step through them,
# changing to each Alerts directory,
# and touching the default.alrt file.
set -A key_sites_2 `list_sites.ksh`
for key_site in ${key_sites_2[*]}
do
 cd /$HCIROOT/$key_site
 cd Alerts
 touch default.alrt
 sleep 5
done

I see there is a sleep in the script and don’t know why so don’t ask, sorry.

Russ Ross
RussRoss318@gmail.com

[Manny Zachodin]

Thank you all for your suggestions. I will try them out.

[Nate Kruse]

Here is our action:  ../../../contrib/clover_email.ksh “%A” “clover@hospitalname.org“

Look in your ../../integrator/contrib directory.  We use two files regarding email – clover_email.ksh and clover_email_alert.tcl

[Jennifer Hardesty]

Russ Ross or anyone who uses this method of touching the default.alrt files:

I have 3 questions.

1.

[Russ Ross]

My response are in green

Russ Ross or anyone who uses this method of touching the default.alrt files:

I have 3 questions.

1.  When you reset the alert monitor, do the time counters on the alerts start counting again or do they automatically fire off?  For example, if I have an alert that is supposed to fire when no new messages have arrived in 20 minutes and this was true when the default.alrt file is reset, will it automatically fire another alert immediately or will it wait 20 more minutes?

I’ve not paid that much attention to that level of detail and don’t know for sure.  

2. Every time you reset the log, how much effect/noise on the engine logs is there?

I don’t reset the log I just touch the default.alrt file.

3.  Have there been any negative effects?

There have been no significant negative effects other than the alert will continue to fire once an hour until the problem is resolved or we toggle off the alert.

Russ Ross
RussRoss318@gmail.com

[Jennifer Hardesty]

Quote:

I don’t reset the log I just touch the default.alrt file.

One of my fellow integration teammates has expressed concern that touching the default.alrt file causes it to be reloaded by the engine every time (hour?), which will result in a multitude of messages filling up the engine log file, making it difficult to troubleshoot any issues we might have with the engine.  

How does Cloverleaf know that default.alrt has been touched?  Is it constantly checking the timestamp?

Thanks again!

[Jennifer Hardesty]

In case anyone is searching for answers later…the entire default.alrt file does in fact reload every time the touch command is used and each alert in the file is repeated in the hcimonitord.log file every time default.alrt is updated.

[Author]

Replies