Add monitoring #36
Add monitoring #36
Conversation
Signed-off-by: Dawid Lesiak <[email protected]>
Signed-off-by: Dawid Lesiak <[email protected]>
for more information, see https://pre-commit.ci
Signed-off-by: Dawid Lesiak <[email protected]>
dawidlesiak
force-pushed
the
add-monitoring
branch
from
c925200 to
8f0ad67
Compare
| /usr/lib/zabbix/externalscripts | ||
| ``` | ||
|
|
||
| > The default location of externalscripts can be changed in `/etc/zabbix/zabbix_server.conf` by editing `ExternalScripts=/usr/lib/zabbix/externalscripts` (make sure to remove #) |
There was a problem hiding this comment.
@dawidlesiak Please use pre-commit locally. See failures: https://results.pre-commit.ci/run/github/649846685/1716814016.AtAIM8RRQYqQuw086aFAZQ
| ## Create the item in Zabbix | ||
| > | ||
| > Item is a tool that gathers data from device, system or external script | ||
| 2. Navigate to **Configuration** > **Hosts**. |
There was a problem hiding this comment.
This numbering makes no sense. Either use 1 everywhere (and let the markdown renderer handle it), or use incremental numbering in each section manually.
| > It's likely that the last used tab is overwriting session data. | ||
|
|
||
| ## Create the item in Zabbix | ||
| > |
| [list of macros](https://www.zabbix.com/documentation/current/en/manual/appendix/macros/supported_by_location#host-inventory) | ||
|
|
||
| ## Create the trigger | ||
| > |
|
|
||
| ## Create the trigger | ||
| > | ||
| > Trigger is a rule that watches for specific events, like high CPU usage. |
| + Item: Select the item for which you want to create a trigger. | ||
| + Function: Defines how your trigger will be activated. | ||
| + Result: Actual rule of activation. | ||
| > If your script returns numeric values directly and/or you don't won't to calculate the average result, simply use `last()` function |
| 4. Test manually. | ||
| - login as **zabbix** user and go to the script location. | ||
| - run your script as zabbix. | ||
| > You can directly incorporate information from Zabbix macros into your command, such as ./script.sh '192.168.10.0'. |
| ## Testing | ||
|
|
||
| 4. Test manually. | ||
| - login as **zabbix** user and go to the script location. |
| done < "$stolen" | ||
|
|
||
| deactivate | ||
|
|
| if grep -qw "$ip" "$fail"; then | ||
| echo 0 | ||
| elif grep -qw "$ip" "$used"; then | ||
| echo 1 |
There was a problem hiding this comment.
use consistent indentation
| @@ -0,0 +1,15 @@ | |||
| #!/bin/bash | |||
There was a problem hiding this comment.
do we actually use flash_read.sh and flash_write.sh files in monitoring process?
There was a problem hiding this comment.
Yes, flash_read.sh and flash_write.sh are executed by Zabbix and are separated into two files for ease of use with zabbix. Meanwhile flash_probe.sh is executed by a crontab or another automation tool due to Zabbix's timeout limitations.
| check_snipeit() { | ||
| ip=$1 | ||
| if [[ "$ip" =~ [0-9] ]]; then | ||
| if osfv_cli snipeit list_used | grep -qw "$ip"; then |
There was a problem hiding this comment.
This is probably not needed, see below.
| if osfv_cli snipeit list_used | grep -qw "$ip"; then | ||
| return 0 # IP is checked out | ||
| else | ||
| osfv_cli snipeit check_out --rte_ip "$ip" |
There was a problem hiding this comment.
Simply try to check out. If platform is used, you will simply get an arror:
Error checking out asset 52
Response data: {'status': 'error', 'messages': 'That asset is not available for checkout!', 'payload': {'asset': '00052'}}
| output_pass="/usr/lib/zabbix/externalscripts/output_pass.txt" | ||
| output_fail="/usr/lib/zabbix/externalscripts/output_fail.txt" | ||
| output_used="/usr/lib/zabbix/externalscripts/output_used.txt" | ||
| stolen="/usr/lib/zabbix/externalscripts/stolen.txt" |
There was a problem hiding this comment.
what is the purpose of this file? Can you briefly document each of them as a comment above?
| output_used="/usr/lib/zabbix/externalscripts/output_used.txt" | ||
| stolen="/usr/lib/zabbix/externalscripts/stolen.txt" | ||
|
|
||
| > "$output_used" |
There was a problem hiding this comment.
When and where do you clear those files? Why not do it here at the beginning of the script?
There was a problem hiding this comment.
i used redirects to clear the file out of habbit, already changed them to rm
> "$output_used"
> "$output_pass"
> "$output_fail"
> "$stolen"
|
|
||
| source /usr/lib/zabbix/externalscripts/venv/bin/activate | ||
|
|
||
| input_file="/usr/lib/zabbix/externalscripts/macros.txt" |
There was a problem hiding this comment.
What is this file, and what it contains?
There was a problem hiding this comment.
$input_file in this case macros.txt
contains list of IPs to use
I explained it in new commit
| ip=$1 | ||
| if [[ "$ip" =~ [0-9] ]]; then | ||
| if osfv_cli snipeit list_used | grep -qw "$ip"; then | ||
| return 0 # IP is checked out |
There was a problem hiding this comment.
This logic seems reversed. When device is taken, I would treat it as some form of error. And we return 0 on sucess, and non-zero on error.
| else | ||
| osfv_cli snipeit check_out --rte_ip "$ip" | ||
| echo "$ip" >> "$stolen" | ||
| return 1 # IP not checked out |
| return 1 # IP not checked out | ||
| fi | ||
| else | ||
| return 1 # IP contains no numbers |
There was a problem hiding this comment.
Return non-zero on error. ideally, different error code for different type of error (not 1 in two places).
| fi | ||
| done < "$input_file" | ||
|
|
||
| while IFS= read -r ip || [ -n "$ip" ]; do |
There was a problem hiding this comment.
Was it added just recently, and should fix the devices being reserved on robot for too long?
| # echo "$ip" >> "$output_used" | ||
| else | ||
| echo "$ip not found in snipeit list_used, probing flash..." | ||
| probe_flash "$ip" |
There was a problem hiding this comment.
How about simply invoking check_in below? After we call probe_flash, we no longer need the device to be checked out, right? This makes another loop and stolen.txt not needed.
Signed-off-by: Dawid Lesiak <[email protected]>
Signed-off-by: Dawid Lesiak <[email protected]>
No description provided.