RHEL 7.1.1503

You need something from me?

Hello Nitin,

Thank you for registering this issue. Here are some additional questions:

Which Zabbix version has been used?

What is the detailed description when you press on [i]?

What is the busyness of Zabbix alerter process? This can be observed by clicking on Monitoring -> Graphs -> Select group: Zabbix servers, -> select host: Zabbix server, -> "Zabbix Internal process busy". Please attach the graph to represent a time period for last 1 day.

Version :- 4.0.1

Last 1 day graph

Thank you for the screens.

Also please attach the graph "Zabbix data gathering process".

How many 'StartPollers=' and 'Timeout=' you have configured in 'zabbix_server.conf'?

Timeout=4

StartPollers="It is commented out, Might be picking default"

StartPollersUnreachable=100 

100 unreachable pollers still having such a high utilisation indicates serious problems in the environment. Most likely, this is is a support request, thus would be a better fit on IRC or other support channels, listed at https://zabbix.org/wiki/Getting_help .

Still, it would be interesting to see how many alerts in which status you have. What does the following return?

select status,count(*) from alerts group by status;

----------------+

status

count

----------------+

----------------+

Let me know if anything else is needed

Assuming default 3 alerters are running then one of them is constantly stuck at 100% busy. (see constant 33.33% alerter load)

Some weird Email media type settings could have caused that. Administration -> Media types -> Email . Check both tabs in config there. 

Nope! Wierd media types were jabber and sms, they were not used anywhere!!

Disabled them anyway!!

Any Update??

In creased the alerters to 50 and escalators to 50 as well. Mails are getting delayed by 1-2 days

Not the SMS and Jabber you were asked about as in the screenshots it is clear the Email is the media used. What about settings in them?

Knowing that you now have 50 alerters and escalators will not help if your mailserver is the bottleneck in the first place. 

No Mailq found on the mail server too!

select status,count(*) from alerts group by status;

This reported:

which means:

There is a lot of messages to deliver.

I will mark and close this ticket as Commercial support required.

Resolved it!!

Seems like problem was with only 1 connection to smtp server. Increased the Concurrent sessions to fix it.