Discussion thread for official Zabbix Template Apache

[AlexL]

This thread is designed to provide grounds for discussion of the upcoming official Zabbix Template for Apache.
The template and details of the template are available in GIT repository.
HTTP Monitoring Template:
Agent Monitoring Template:

Zabbix is always looking for ways to improve our services and to make our users happier.
We pride ourselves on doing our best each and every day, but we know that there is always something more to learn.
We would like to hear back from you to know what have you liked and what would you improve in the template.

[AlexL]

Templates have been released and are available under the following URL:
HTTP Monitoring Template: https://git.zabbix.com/projects/ZBX/...pp/apache_http
Agent Monitoring Template: https://git.zabbix.com/projects/ZBX/...p/apache_agent

[multazim]

Looks like these templates are not compatible with Zabbix 4.0.
It would be great if official templates are released for 4.0 as well.

[gospodin.horoshiy]

Hi! This template relies on Javascript so porting it to 4.0 is not that easy. While not that complete, you can try this template instead for 4.0 https://github.com/v-zhuravlev/zbx_template_apache2

[langemol]

I can't seem to get half of the items working. The status page returns

 [TABLE]
[TR]
[TD] 			
 {"Date":"Fri, 01 Nov 2019 10:48:44 GMT","Server":"Apache/2.4.29 (Ubuntu)","Location":"https://127.0.0.1/server-status?auto","Length":220,"Type":"text/html; charset=iso-8859-1","Workers":{"waiting":0,"starting":0,"reading":0,"sending":0,"keepalive":0,"dnslookup":0,"closing":0,"logging":0,"finishing":0,"cleanup":0,"slot":0}} 			[/TD]
 		[/TR]
[/TABLE]

Half of the info the template needs seems to be missing on the page.
I checked `/etc/apache2/mods-enabled/status.conf` and it has `ExtendedStatus On`.
I'm running on Ubuntu 18.04

[jgiucci]

Hi! Great template, nice way to get an example on how to use javascript in preprocessing. Got it working after enabling mod_status on Apache.

The only thing I had to modify is the first preprocessing step for the "Apache: version" item. I changed the parameter to ":Server" from ":ServerVersion", since that is how it is being reported:

{"Date":"Fri, 22 Nov 2019 18:12:55 GMT","Server":"Apache/2.4.6 (Red Hat Enterprise Linux) PHP/5.4.16","Length":429,"Type":... Don't know if it is just for this version of Apache.

Saludos!

[unficyp]

How can i set http scheme ? Seems like Agent is always connecting using https.
thanks

[jkangas]

Curious why there is no default Trigger for Apache Slots being full. Also curious how to prevent the server-status Items from going into a disabled state when all apache slots are full. We just applied this template and we're running into this exact issue. Http call to get server-status timesout due to no apache slots and then the server-status items become disabled as a result. How would we prevent this from happening?

[Yasumi]

There is a part that I do not understand the specifications of this template so much.

Below is the information I got with "Get Apache".
{"Connection":"close","Length":211,"Type":"text/html; charset=iso-8859-1","Date":"Thu, 03 Sep 2020 03:54:52 GMT","Server":"Apache/2.4.37 (centos) OpenSSL/1.1.1c","Workers":{"waiting":0,"starting":0,"readi ng":0,"sending":0,"keepalive":0,"dnslookup":0,"c lo sing":0,"logging":0,"finishing":0,"cleanup":0,"slo t":0}}

Apparently, it seems that the information corresponding to the following items has not been acquired.
Is this normal behavior?
apache.workers_total.busy
apache.workers_total.idle
apache.requests.rate
apache.requests
apache.bytes.rate
apache.bytes

[Leonid Kabanov]

Type of "Apache: Total bytes" should be changed to "Numeric (unsigned)":

6549:20210302:000202.127 item "web17 server:apache.bytes" became not supported: Value 1000003696640 is too small or too large.
6550:20210302:000302.469 error reason for "web17 server:apache.bytes" changed: Value 1000005673984 is too small or too large.

[Lizzardd]

I'm a little confused as to the trigger below

Apache: Process is not running

-

{TEMPLATE_NAMEroc.num["{$APACHE.PROCESS_NAME}"].last()}=0

This will always be in "Problem" state while apache is running as the results will be greater than 0.
When Apache is stopped this will be in state "OK"

I see there are also dependencies off of this trigger, is this the correct and intended behaviour?

[Alec]

Hello,

Since the update to zabbix 5.2.5 the "zabbix agent 2" on the servers where this template is activated crash :

zabbix_agent2[488]: panic: runtime error: index out of range [0] with length 0
zabbix_agent2[488]: goroutine 45076 [running]:
zabbix_agent2[488]: zabbix.com/plugins/proc.getProcesses(0xb, 0xc0003c3400, 0x79, 0x80, 0x0, 0x0)
zabbix_agent2[488]: /build/zabbix/src/zabbix-5.2.5/src/go/plugins/proc/procfs_linux.go:142 +0x93f
zabbix_agent2[488]: zabbix.com/plugins/proc.(*Plugin).Collect(0x124f220, 0x0, 0x0)
zabbix_agent2[488]: /build/zabbix/src/zabbix-5.2.5/src/go/plugins/proc/proc_linux.go:217 +0xb7
zabbix_agent2[488]: zabbix.com/internal/agent/scheduler.(*collectorTask).perform.func1(0xc000197 a40, 0xe45218, 0xc0002280a0)
zabbix_agent2[488]: /build/zabbix/src/zabbix-5.2.5/src/go/internal/agent/scheduler/task.go:104 +0x6f
zabbix_agent2[488]: created by zabbix.com/internal/agent/scheduler.(*collectorTask).perform
zabbix_agent2[488]: /build/zabbix/src/zabbix-5.2.5/src/go/internal/agent/scheduler/task.go:102 +0xb1

[bes]

https://support.zabbix.com/browse/ZBX-19209

[ddougan]

I'm seeing these errors in the zabbix_server log (for Apache on the server itself):

1. Failed: cannot extract value from json by path "$.IdleWorkers": no data matches the specified path
2987:20210424:104340.164 error reason for "Zabbix server:apache.workers_total.busy" changed: Preprocessing failed for: {"Date":"Sat, 24 Apr 2021 17:43:40 GMT","Server":"Apache/2.4.37 (centos) OpenSSL/1.1.1g","Length"...
1. Failed: cannot extract value from json by path "$.BusyWorkers": no data matches the specified path
2987:20210424:104340.164 error reason for "Zabbix server:apache.bytes.rate" changed: Preprocessing failed for: {"Date":"Sat, 24 Apr 2021 17:43:40 GMT","Server":"Apache/2.4.37 (centos) OpenSSL/1.1.1g","Length"...
1. Failed: cannot extract value from json by path "$["Total kBytes"]": no data matches the specified path

This is on Zabbix 5.2 on CentOS 8, updated recently. I'm not sure what these are being caused by.

[Colttt]

Hello,

I got the error:

HTTP get error: cannot resolve [http://127.0.0.1:60080/server-status?auto]

but if I run this command on the machine it works fine:

curl http://127.0.0.1:60080/server-status?auto
127.0.0.1
ServerVersion: Apache/2.4.38 (Debian) OpenSSL/1.1.1d
ServerMPM: prefork
Server Built: 2020-08-25T20:08:29
CurrentTime: Monday, 20-Sep-2021 16:57:17 CEST
RestartTime: Wednesday, 08-Sep-2021 09:43:58 CEST
ParentServerConfigGeneration: 14
ParentServerMPMGeneration: 13
ServerUptimeSeconds: 1062799
ServerUptime: 12 days 7 hours 13 minutes 19 seconds
Load1: 0.00
Load5: 0.00
Load15: 0.00
Total Accesses: 64530
Total kBytes: 1252525
Total Duration: 777144864
CPUUser: 17.49
CPUSystem: 37.45
CPUChildrenUser: 387.39
CPUChildrenSystem: 113.01
CPULoad: .0522526
Uptime: 1062799
ReqPerSec: .060717
BytesPerSec: 1206.8
BytesPerReq: 19875.8
DurationPerReq: 12043.2
BusyWorkers: 1
IdleWorkers: 9
Scoreboard: __...__._.._..._..W......_...._................... .................................................. ..................................................
TLSSessionCacheStatus
CacheType: SHMCB
CacheSharedMemory: 512000
CacheCurrentEntries: 11
CacheSubcaches: 32
CacheIndexesPerSubcaches: 88
CacheTimeLeftOldestAvg: 67
CacheTimeLeftOldestMin: 8
CacheTimeLeftOldestMax: 93
CacheIndexUsage: 0%
CacheUsage: 0%
CacheStoreCount: 1672
CacheReplaceCount: 0
CacheExpireCount: 1652
CacheDiscardCount: 0
CacheRetrieveHitCount: 1
CacheRetrieveMissCount: 5
CacheRemoveHitCount: 9
CacheRemoveMissCount: 7

any hints for this?

[Daniel4711]

Hey, we just noticed that the command from the template does not work with the old agent.

Old Agent : zabbix_get -s Agent1HostIP -k 'web.page.get["127.0.0.1","/server-status","80"]'
New Agent2: zabbix_get -s Agent2HostIP -k 'web.page.get["http://127.0.0.1:80/server-status"]'

The "Old Agent" command is working on Agent2.
Maybe this will help you

I got the info from this thread:

https://www.zabbix.com/forum/zabbix-help/408374-apache-by-agent-cannot-resolve?p=412317#post412317

[nikolam]

Hello,

I have upgraded Zabbix from 5.4 LTS to 6.0 LTS version and wanted to use apache2 template check that is working with new Zabbix version.

I've edited in template process name to {$APACHE.PROCESS_NAME} from httpd to apache2 and port and status scheme to 443 and https.

One server is working alright, while others aren't. Reason is the IP address. On that server where is status check working, it has IP redirection to domain name.
Just wanted to write in case someone had same issues as me trying to make it work.

Is there a way for zabbix to use domain name as status check? Macro that taking value of agent domain name and trying to do /server-status/ .
If its like this it would work no matter what for everyone, due that many servers don't have IP redirect to the domain name where I put /server-status/ check.

If someone has a solution on their mind let me know or if my idea for macro is good then what is the macro that grabs that value?

TIA