[ZBX-16441] Huge queue, gap on graphs Created: 2019 Jul 31  Updated: 2019 Oct 29  Resolved: 2019 Oct 10

Status: Closed
Project: ZABBIX BUGS AND ISSUES
Component/s: None
Affects Version/s: 4.0.11
Fix Version/s: None

Type: Problem report Priority: Critical
Reporter: Kirill Varnakov Assignee: Edgars Melveris
Resolution: Fixed Votes: 0
Labels: gap, graphs, queue
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Attachments: PNG File Graphs.png     PNG File Latest data.png     PNG File Perfomance after problem.png     PNG File Perfomance before problem.png     PNG File Perfomance.png     PNG File Preprocessing queue.png     PNG File Proxy best resolution.png     PNG File Proxy perfomance after problem.png     PNG File Proxy perfomance.png     PNG File Queue.png    

 Description   

Hi. Something strange happens. After several days of work, appears gap on graphs and huge queue. But nothing criminal in logs and performance of server.

I enabled debug on server and proxy, checked DB tables and logs, increased virtual machine perfomance. It didn't help.

Maybe you know, where to dig?



 Comments   
Comment by Glebs Ivanovskis [ 2019 Jul 31 ]

Do you have a graph of preprocessing queue?

Right after 14:15 "history index cache, % free" jumps from it's normal value to 100%, which means that there are virtually no items left in history cache (all were successfully written to DB, I guess?) Honestly, I haven't seen graphs like this one before, but there might be a buildup of values before history cache — in preprocessing.

Comment by Kirill Varnakov [ 2019 Aug 01 ]

Added new screenshots. No problems... My brain feels bad )

Comment by Kirill Varnakov [ 2019 Aug 01 ]

In my installation, it happens with hosts from only one main proxy, but proxy has no problem too.

Comment by Glebs Ivanovskis [ 2019 Aug 01 ]

Would be nice to get proxy graphs in better resolution around 14:00 on July 31 (when the problem started on server side). There is something going on, but it's hard to tell, because data is likely from trends at this scale.

Comment by Kirill Varnakov [ 2019 Aug 01 ]

Yes, surely. In attach.

Comment by Glebs Ivanovskis [ 2019 Aug 01 ]

This is a passive proxy, right?

Comment by Kirill Varnakov [ 2019 Aug 01 ]

Yes. Another one in active mode with no problems.

Comment by Glebs Ivanovskis [ 2019 Aug 01 ]

Are you still experiencing the issue or is it all over? Could you show server graphs when it works as usual (roughly at the same scale as Perfomance.png)?

Comment by Kirill Varnakov [ 2019 Aug 01 ]

It still exists. 

Comment by Kirill Varnakov [ 2019 Aug 01 ]

Now, problem has disappeared. After restart maybe or work of housekeeper... It happens during 3 weeks. But what I do wrong? 

Comment by Glebs Ivanovskis [ 2019 Aug 01 ]

After restart of what? How does it look on the proxy side, the moment when the issue disappeared?

I have no idea what it could be, sorry...

Comment by Kirill Varnakov [ 2019 Aug 01 ]

After server restart, but it connected with end of housekeeper work. Proxy perfomance after problem attached.

Comment by Glebs Ivanovskis [ 2019 Aug 01 ]

After server restart

Oh, I see it now. It's barely visible.

It happens during 3 weeks.

Does the issue happen regularly?

Comment by Kirill Varnakov [ 2019 Aug 01 ]

Yes, waiting for the next coming...

Comment by Kirill Varnakov [ 2019 Oct 10 ]

After updating to 4.0.12 the problem disappeared.

Generated at Fri Apr 04 11:49:49 EEST 2025 using Jira 9.12.4#9120004-sha1:625303b708afdb767e17cb2838290c41888e9ff0.