[ZBX-14405] Zabbix - Web Scenarios issue Created: 2018 May 28  Updated: 2018 Jun 05  Resolved: 2018 Jun 05

Status: Closed
Project: ZABBIX BUGS AND ISSUES
Component/s: Appliance (L)
Affects Version/s: 3.4.8
Fix Version/s: None

Type: Patch request Priority: Major
Reporter: Fimmick Assignee: Unassigned
Resolution: Won't fix Votes: 0
Labels: webmonitoring
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Attachments: Text File result.txt     PNG File zabbix.png    

 Description   

The Zabbix server is working around two months, everything is working other than the web scenarios monitoring. The response time of those monitoring URLs in different server are very strange, some response time are same around 5.xx second, but we found nothing error in the log and confirmed that there is no cronjob is running.

Result:
See screenshot



 Comments   
Comment by Alexey Pustovalov [ 2018 May 28 ]

I recommend you increase "http poller" process debug level to 5, then check Zabbix server log file when the issue happens again.

Comment by Fimmick [ 2018 May 28 ]

Thanks Alexey,

I have enabled the debug log to level 5, we compared with some normal case, the logging is no different other than the response time and speed. Could you please advice the next step.

result.txt

Comment by Alexey Pustovalov [ 2018 May 28 ]

why? I see 5 seconds response:

  6607:20180528:104300.474 process_httptest() go to URL "https://xxxxxx/xxxx/xxxx/xx/"
  6607:20180528:104306.179 process_httptest() page.data from https://xxxxxx/xxxx/xxxx/xx/
Comment by Fimmick [ 2018 May 28 ]

yes, it's the problem, but the debug log can't trace the root cause. That's why we asks for another ways to trace it?

Comment by Alexey Pustovalov [ 2018 May 28 ]

the root cause you need to check on remote site. It is just slow response.

Comment by Fimmick [ 2018 May 28 ]

Actually, most of the time, the response time of the monitoring url is very fast which is around 200 ms.
The most strange thing is the slow response time is almost all exactly in 5.7s.

If I want to check in my remote site (agent), any debug level of zabbix_agentd.log can change in order to figure out the problem? Please advise. Thanks!

Comment by Alexey Pustovalov [ 2018 May 28 ]

web-scnerios are executed on Zabbix server (or proxy) level, Zabbix agent is not involved to web-scenario checks.

Comment by Fimmick [ 2018 Jun 05 ]

Still not fixed yet...any idea? Thanks!

Comment by Alexey Pustovalov [ 2018 Jun 05 ]

I recommend you check network level + web-application level. Zabbix just reports about latency, it is not inside of Zabbix server/proxy.

Comment by Fimmick [ 2018 Jun 05 ]

Checked the access log of my client server, the receiving time of web scenario access request is:
XXX.XXX.XXX.XXX - - [28/May/2018:10:43:06 +0800] "GET /xxxxxx/xxxx/xxxx/xx/ HTTP/1.1" 200 55790

it's meaning that there is around 5s gap in between, I wonder if there is any process will run in the zabbix server after logging below record, and before send out to the client server?
6607:20180528:104300.474 process_httptest() go to URL "https://xxxxxx/xxxx/xxxx/xx/"

How could I determine if 100% is network issue?

Comment by Alexey Pustovalov [ 2018 Jun 05 ]

Try to check PING, latency and other network metrics. Actually it is better if you ask the question on our forum (https://www.zabbix.com/forum/). JIRA is designated only for bug reports, but I do not see any issues in Zabbix in such case.

Generated at Mon May 05 07:13:51 EEST 2025 using Jira 9.12.4#9120004-sha1:625303b708afdb767e17cb2838290c41888e9ff0.