Details
-
Bug
-
Resolution: Unresolved
-
Low
-
None
-
9.12.0, 9.11.3, 9.14.0
-
9.11
-
1
-
Severity 3 - Minor
-
Description
Issue Summary
OutgoingMailServerConnectionIpdJob in 'In product diagnostics' (IPD) runs every minute and sends EHLO requests to the SMTP server, regardless if Jira is sending email or regardless of the what the interval for the MailQueueService is.
The issue is multiplied by the number of nodes in a cluster as each node would be sending such EHLO requests to the mail server, thus, making it look like a spam attack or causing excessive logging overhead on the mail servers
This is reproducible on Data Center: Yes
Steps to Reproduce
- Deploy a Jira DC 9.12 or above - 2 nodes will do
- Add -Dmail.debug=true to the startup parameters, restart the nodes for the parameter to take effect
- add com.atlassian.jira.ipd as a logging package with increased DEBUG log level under logging and profiling in Jira
- configure SMTP server
- monitor catalina.out file for the occurences of EHLO messages (e.g. tail -f catalina.out | grep EHLO) on each node
Expected Results
- EHLO messages would at least only happen from one node only (i.e. it should be coordinated through clusteredjob or another mechanism) and at a less frequent interval
- Ideally the IPD or OutgoingMailServerConnectionIpdJob interval should be configurable
Actual Results
Jira sends EHLO requests to the SMTP server every minute from each node, these would be evident in catalina.out:
EHLO ip-Jira-Node-IP-address
EHLO ip-Jira-Node-IP-address
The following log entries will appear in atlassian-jira.log at the same time:
2024-03-19 14:42:41,938+1100 ipd-worker:thread-1 DEBUG [c.a.d.internal.ipd.DefaultIpdJobRunner] Running job com.atlassian.jira.ipd.mailqueue.MailQueueJobRunningInProductDiagnosticJob
2024-03-19 14:42:41,939+1100 ipd-worker:thread-1 DEBUG [c.a.d.internal.ipd.DefaultIpdJobRunner] Running job com.atlassian.jira.ipd.db.DatabaseConnectionStateInProductDiagnosticJob
2024-03-19 14:42:41,939+1100 ipd-worker:thread-1 DEBUG [c.a.d.internal.ipd.DefaultIpdJobRunner] Running job com.atlassian.jira.ipd.db.DatabaseLatencyInProductDiagnosticJob
2024-03-19 14:42:41,940+1100 ipd-worker:thread-1 DEBUG [c.a.d.internal.ipd.DefaultIpdJobRunner] Running job com.atlassian.jira.ipd.mailqueue.MailQueueItemsJob
2024-03-19 14:42:41,940+1100 ipd-worker:thread-1 DEBUG [c.a.d.internal.ipd.DefaultIpdJobRunner] Running job com.atlassian.jira.ipd.mail.OutgoingMailServerConnectionIpdJob
Workaround
Apart from disabling the IPD completely via Jira UI - there is no other workaround available at the moment
Attachments
Issue Links
- mentioned in
-
Page Loading...