How can create a custom item for a script

[bab]

I have a script that return follow output :

 |spa       |dae_0_1_disk_12  |    4771
                    |spa       |dae_0_1_disk_14  |    4916
                    |spa       |dae_0_1_disk_16  |   337.5
                    |spa       |dae_0_1_disk_18  |   91.52
                    |spa       |dae_0_1_disk_2   |    6375
                    |spa       |dae_0_1_disk_20  |    9826
                    |spa       |dae_0_1_disk_22  |   10133
                    |spa       |dae_0_1_disk_24  |   127.1
                    |spa       |dae_0_1_disk_4   |   194.9
                    |spa       |dae_0_1_disk_6   |   146.0
                    |spa       |dae_0_1_disk_8   |   11435
                    |spa       |dae_0_2_disk_0   |   15726
                    |spa       |dae_0_2_disk_10  |   14060
                    |spa       |dae_0_2_disk_12  |   15737
                    |spa       |dae_0_2_disk_14  |   11848
                    |spa       |dae_0_2_disk_16  |    5843
                    |spa       |dae_0_2_disk_18  |   15882
                    |spa       |dae_0_2_disk_2   |   15678
                    |spa       |dae_0_2_disk_20  |   549.0
                    |spa       |dae_0_2_disk_22  |   346.4
                    |spa       |dae_0_2_disk_24  |   776.8
                    |spa       |dae_0_2_disk_4   |   865.2
                    |spa       |dae_0_2_disk_6   |    9026
                    |spa       |dae_0_2_disk_8   |    4180
                    |spa       |dae_0_3_disk_0   |    6810
​

As the output will change every 5 minutes (here is "4916 , 337.5 , 91.52, .........".Is that possible create items for this output ? For example it shows
dae_0_1_disk_14 = 4916
dae_0_1_disk_16 = 337.5
dae_0_1_disk_18 = 91.52
.
.
.
.

[Hamardaban]

Of course it is possible - use dependent item with a regex type preprocessing.

https://www.zabbix.com/documentation/6.0/en/manual/config/items/itemtypes/dependent_items

If the data set can change and you don't want to create many identical elements with your hands, use LLD.

https://www.zabbix.com/documentation/6.0/en/manual/discovery/low_level_discovery

For details, refer to the documentation.

[bab]

Of course it is possible - use dependent item with a regex type preprocessing.
https://www.zabbix.com/documentation...ependent_items

If the data set can change and you don't want to create many identical elements with your hands, use LLD.
https://www.zabbix.com/documentation...evel_discovery

For details, refer to the documentation.

Thanks. Actually I prefer to use LLD but I am just running follow command on zabbix server :

"uemcli -d 172.2.2.2 -u USER -securePassword -sslPolicy store /metrics/value/rt -path sp..physical.disk..responseTime show "

and it returns such as this

|spa |dae_0_1_disk_14 | 4916
|spa |dae_0_1_disk_16 | 337.5
|spa |dae_0_1_disk_18 | 91.52
|spa |dae_0_1_disk_2 | 6375
|spa |dae_0_1_disk_20 | 9826
|spa |dae_0_1_disk_22 | 10133​

Would you please some help me how can create this LLD

[cyber]

This output is not the best to convert, but doable... but you could try to obtain data already as CSV ( at least https://www.dell.com/support/manuals...0b2&lang=en-us shows a command with -output csv) or even better json (I cannot find anything in docs about this one) ... Then you can skip all the converting in Zabbix side .. With current output format you should do "convert to CSV" and then "CSV to JSON". Then you can point LLD macros to correct place in JSON. and then you can do all your discovery stuff...
If you can format data already at origin point properly, yu can skip some conversion steps. Rest of it is in the link given above...

[bab]

This output is not the best to convert, but doable... but you could try to obtain data already as CSV ( at least https://www.dell.com/support/manuals...amp;lang=en-us shows a command with -output csv) or even better json (I cannot find anything in docs about this one) ... Then you can skip all the converting in Zabbix side .. With current output format you should do "convert to CSV" and then "CSV to JSON". Then you can point LLD macros to correct place in JSON. and then you can do all your discovery stuff...
If you can format data already at origin point properly, yu can skip some conversion steps. Rest of it is in the link given above...

I could convert output such as follow :

Timestamp,SP,disk,Average Response Time us
2023-01-10 05:33:00,spa,dae_0_1_disk_0,730.1
2023-01-10 05:33:00,spa,dae_0_1_disk_10,159.5
2023-01-10 05:33:00,spa,dae_0_1_disk_12,20786
2023-01-10 05:33:00,spa,dae_0_1_disk_14,9578
2023-01-10 05:33:00,spa,dae_0_1_disk_16,20647
2023-01-10 05:33:00,spa,dae_0_1_disk_18,20.97
2023-01-10 05:33:00,spa,dae_0_1_disk_2,4902
2023-01-10 05:33:00,spa,dae_0_1_disk_20,19.72
2023-01-10 05:33:00,spa,dae_0_1_disk_22,9451
2023-01-10 05:33:00,spa,dae_0_1_disk_24,19.11
2023-01-10 05:33:00,spa,dae_0_1_disk_4,23060
2023-01-10 05:33:00,spa,dae_0_1_disk_6,986.3
2023-01-10 05:33:00,spa,dae_0_1_disk_8,314.4
2023-01-10 05:33:00,spa,dae_0_2_disk_0,130.0
2023-01-10 05:33:00,spa,dae_0_2_disk_10,18438
2023-01-10 05:33:00,spa,dae_0_2_disk_12,498.4
2023-01-10 05:33:00,spa,dae_0_2_disk_14,27105
2023-01-10 05:33:00,spa,dae_0_2_disk_16,630.4
2023-01-10 05:33:00,spa,dae_0_2_disk_18,177.4
2023-01-10 05:33:00,spa,dae_0_2_disk_2,174.6
2023-01-10 05:33:00,spa,dae_0_2_disk_20,7830
2023-01-10 05:33:00,spa,dae_0_2_disk_22,193.0
2023-01-10 05:33:00,spa,dae_0_2_disk_24,1783
​

What should I do now ?

[Hamardaban]

With current output format you should do (1) "convert to CSV" and then (2) "CSV to JSON". Then you can (3) point LLD macros to correct place in JSON . and then you can do (4) all your discovery stuff...

Step (1) - Done!
Step (2) - see https://www.zabbix.com/documentation...ng/csv_to_json
Step (3) - https://www.zabbix.com/documentation...#custom-macros
Step (4) - https://www.zabbix.com/documentation...tem_prototypes

[cyber]

So I assume your item works and retrieves data? Add a preprocessing to it, "CSV to JSON". then your data is saved as

[
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_0","Aver age Response Time us":"730.1"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_10","Ave rage Response Time us":"159.5"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_12","Ave rage Response Time us":"20786"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_14","Ave rage Response Time us":"9578"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_16","Ave rage Response Time us":"20647"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_18","Ave rage Response Time us":"20.97"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_2","Aver age Response Time us":"4902"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_20","Ave rage Response Time us":"19.72"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_22","Ave rage Response Time us":"9451"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_24","Ave rage Response Time us":"19.11"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_4","Aver age Response Time us":"23060"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_6","Aver age Response Time us":"986.3"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_1_disk_8","Aver age Response Time us":"314.4"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_2_disk_0","Aver age Response Time us":"130.0"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_2_disk_10","Ave rage Response Time us":"18438"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_2_disk_12","Ave rage Response Time us":"498.4"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_2_disk_14","Ave rage Response Time us":"27105"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_2_disk_16","Ave rage Response Time us":"630.4"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_2_disk_18","Ave rage Response Time us":"177.4"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_2_disk_2","Aver age Response Time us":"174.6"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_2_disk_20","Ave rage Response Time us":"7830"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_2_disk_22","Ave rage Response Time us":"193.0"},
{"Timestamp":"2023-01-10 05:33:00","SP":"spa","disk":"dae_0_2_disk_24","Ave rage Response Time us":"1783"}
]

Add a discovery. Type of information -> text. Set its type as "dependent item" and point master item to the item that retrieves your data. Add to LLD macros tab "{#DISK} -> $.disk".
Add item prototype to discovery, Type of information -> float. Set its type as "dependent item" and point master item to the item that retrieves your data. (yes the same as in discovery). Add preprocessing "JSONPath -> $.[?(@.disk=="{#DISK}")].['Average Response Time us'].first()"
EDIT2: name your item prototype as disk_avgerage[{#DISK}] for example... it has to contain LLD macro, but rest is up to you...

Now, when your main item acquires data, first time all the items are created, based on disk name. Second time, they are already populated with a value.

EDIT: oh short answer came earlier..

[Hamardaban]

The full answer is much better! :-)

[bab]

So I assume your item works and retrieves data? Add a preprocessing to it, "CSV to JSON". then your data is saved as

Add a discovery. Type of information -> text. Set its type as "dependent item" and point master item to the item that retrieves your data. Add to LLD macros tab "{#DISK} -> $.disk".
Add item prototype to discovery, Type of information -> float. Set its type as "dependent item" and point master item to the item that retrieves your data. (yes the same as in discovery). Add preprocessing "JSONPath -> $.[?(@.disk=="{#DISK}")].['Average Response Time us'].first()"
EDIT2: name your item prototype as disk_avgerage[{#DISK}] for example... it has to contain LLD macro, but rest is up to you...

Now, when your main item acquires data, first time all the items are created, based on disk name. Second time, they are already populated with a value.

EDIT: oh short answer came earlier..

I have created an item on my server with userparameter such as attach pic 11.jpg and it is returns follow output :

Storage system address: 172.22.30.29
Storage system port: 443
HTTPS connection

Timestamp,SP,disk,Average Response Time us
2023-01-15 15:27:45,spa,dae_0_1_disk_0,2994
2023-01-15 15:27:45,spa,dae_0_1_disk_10,150.5
2023-01-15 15:27:45,spa,dae_0_1_disk_12,19662
2023-01-15 15:27:45,spa,dae_0_1_disk_14,41225
2023-01-15 15:27:45,spa,dae_0_1_disk_16,222.9
2023-01-15 15:27:45,spa,dae_0_1_disk_18,21.00
2023-01-15 15:27:45,spa,dae_0_1_disk_2,12836
2023-01-15 15:27:45,spa,dae_0_1_disk_20,19.57
2023-01-15 15:27:45,spa,dae_0_1_disk_22,37016
2023-01-15 15:27:45,spa,dae_0_1_disk_24,33111
2023-01-15 15:27:45,spa,dae_0_1_disk_4,274.4
2023-01-15 15:27:45,spa,dae_0_1_disk_6,19460
2023-01-15 15:27:45,spa,dae_0_1_disk_8,9598
2023-01-15 15:27:45,spa,dae_0_2_disk_0,5841
2023-01-15 15:27:45,spa,dae_0_2_disk_10,184.8
2023-01-15 15:27:45,spa,dae_0_2_disk_12,209.1
2023-01-15 15:27:45,spa,dae_0_2_disk_14,17660
2023-01-15 15:27:45,spa,dae_0_2_disk_16,20.77
2023-01-15 15:27:45,spa,dae_0_2_disk_18,21181
2023-01-15 15:27:45,spa,dae_0_2_disk_2,8786​

In the next step I have created a discover rule with attached item.prototype and preprocessing but when I click on test it shows error . How can solve it ?

[cyber]

Remove that part first

Storage system address: 172.22.30.29
Storage system port: 443
HTTPS connection​

you only need CSV part...

When you test, give also value to macro ... replace that {#DISK} there with real value from data (like dae_0_2_disk_16), so you can compare, that correct thing will be extracted. I also don't see anything entered into "value" field during testing... ? You need to feed it with proper json...

[bab]

Remove that part first

Storage system address: 172.22.30.29
Storage system port: 443
HTTPS connection​

you only need CSV part...

When you test, give also value to macro ... replace that {#DISK} there with real value from data (like dae_0_2_disk_16), so you can compare, that correct thing will be extracted. I also don't see anything entered into "value" field during testing... ? You need to feed it with proper json...

At the first step I have created an item such as emc1.jpg that returns follow output :

Timestamp,SP,disk,Average Response Time us
2023-01-19 21:45:20,spa,dae_0_1_disk_0,22.62
2023-01-19 21:45:20,spa,dae_0_1_disk_10,49368
2023-01-19 21:45:20,spa,dae_0_1_disk_12,20.83
2023-01-19 21:45:20,spa,dae_0_1_disk_14,19.21
2023-01-19 21:45:20,spa,dae_0_1_disk_16,12239
2023-01-19 21:45:20,spa,dae_0_1_disk_18,49557
2023-01-19 21:45:20,spa,dae_0_1_disk_2,12074
2023-01-19 21:45:20,spa,dae_0_1_disk_20,21.17
2023-01-19 21:45:20,spa,dae_0_1_disk_22,8135​

in the second step I have create a discovery rule such as emc2.jpg
in the next step I have created an itemprototype such as emc3.jpg with preprocessing emc4.jpg
now when click on test it shows error (emc5.jpg)
Also I gave value to macro but returns error (emc6.jpg)
What is the issue ?

[bab]

At the first step I have created an item such as emc1.jpg that returns follow output :

Timestamp,SP,disk,Average Response Time us
2023-01-19 21:45:20,spa,dae_0_1_disk_0,22.62
2023-01-19 21:45:20,spa,dae_0_1_disk_10,49368
2023-01-19 21:45:20,spa,dae_0_1_disk_12,20.83
2023-01-19 21:45:20,spa,dae_0_1_disk_14,19.21
2023-01-19 21:45:20,spa,dae_0_1_disk_16,12239
2023-01-19 21:45:20,spa,dae_0_1_disk_18,49557
2023-01-19 21:45:20,spa,dae_0_1_disk_2,12074
2023-01-19 21:45:20,spa,dae_0_1_disk_20,21.17
2023-01-19 21:45:20,spa,dae_0_1_disk_22,8135​

in the second step I have create a discovery rule such as emc2.jpg
in the next step I have created an itemprototype such as emc3.jpg with preprocessing emc4.jpg
now when click on test it shows error (emc5.jpg)
Also I gave value to macro but returns error (emc6.jpg)
What is the issue ?

Attached emc6.jpg

[cyber]

Did you enter something into "value" filed also, when testing?
I cannot find from pics, where do you apply "csv to json" preprocessing? emc1.jpg does not have any preprocessing steps...

[bab]

Did you enter something into "value" filed also, when testing?
I cannot find from pics, where do you apply "csv to json" preprocessing? emc1.jpg does not have any preprocessing steps...

According to emc6.jpg I just insert parameter in macro part for test
I used CSV to json according emc4.jpg in item prototype in discovery

[cyber]

That pic show only jsonpath preprocessing, but no "csv to json", please add it to the item on pic emc1.jpg....
And if you test, you need to give also starting data to test, not just one value to macro...

I tried to restore my testing template, adding it here... "should" work... but I had no time to test it thoroughly.

[bab]

That pic show only jsonpath preprocessing, but no "csv to json", please add it to the item on pic emc1.jpg....
And if you test, you need to give also starting data to test, not just one value to macro...

I tried to restore my testing template, adding it here... "should" work... but I had no time to test it thoroughly.

Thanks but after add template to that host and select it in discovery role to execute it shows "Cannot send request: wrong discovery rule type."
Also this returns numbers with microsecond unit how can change it to millisecond unit ?