Confluence continuously uses a disk

[Yuriy Gurin]

Confluence: 7.16.0

Running on: Synology 1522+ (self-hosted, in the container)

Database: MySQL 8.2.0

Database plugin: mysql-connector-j-8.1.0.jar

After installation I see that confluence is using my disk almost all the time. iotop says that mysqld is causing the issue. I have toggled it into debug mode and see that confluence is writing some data in the scheduler_run_details table every second. It doesn't look like correct behaviour. What can cause this issue?

mysqld last queries (last 3 seconds):

commit
insert into scheduler_run_details (job_id, start_time, duration, outcome, message, id) values ('SystemMaintenanceTaskQueueFlusherScheduledJob', '2023-12-17 00:01:51.551', 3, 'S', '', 181438776)
SET autocommit=0
SELECT @@session.transaction_read_only
commit
SET autocommit=0
SELECT @@session.transaction_read_only
commit
select journalent0_.entry_id as entry_id1_52_, journalent0_.journal_name as journal_2_52_, journalent0_.creationdate as creation3_52_, journalent0_.type as type4_52_, journalent0_.message as message5_52_ from journalentry journalent0_ where journalent0_.entry_id>0 and journalent0_.journal_name='system_maintenance' and journalent0_.creationdate<='2023-12-17 00:01:50.552' order by journalent0_.entry_id asc limit 20
SET autocommit=0
SELECT @@session.transaction_read_only
commit
SET autocommit=0
SELECT @@session.transaction_read_only
commit
insert into scheduler_run_details (job_id, start_time, duration, outcome, message, id) values ('migration-plugin:mapi-status-consumer-job-id', '2023-12-17 00:01:51.052', 3, 'S', '', 181438775)
SET autocommit=0
SELECT @@session.transaction_read_only
commit
SET autocommit=0
commit
select mapitaskma0_.taskId as taskid1_29_, mapitaskma0_.cloudId as cloudid2_29_, mapitaskma0_.commandName as commandn3_29_, mapitaskma0_.jobId as jobid4_29_, mapitaskma0_.lastUpdate as lastupda5_29_, mapitaskma0_.planId as planid6_29_, mapitaskma0_.status as status7_29_ from MIG_MAPI_TASK_MAPPING mapitaskma0_ where mapitaskma0_.status='CHECKS_IN_PROGRESS' order by mapitaskma0_.lastUpdate
SET autocommit=0
SELECT @@session.transaction_read_only
commit
SET autocommit=0

set session transaction read write
SELECT @@session.transaction_read_only
commit
SET autocommit=0
set session transaction read only
set session transaction read write
set session transaction read write
SELECT @@session.transaction_read_only
SELECT @@session.transaction_read_only
SELECT @@session.transaction_read_only
commit
commit
commit
SET autocommit=0
SET autocommit=0
SET autocommit=0
set session transaction read only
set session transaction read only

SELECT @@session.transaction_read_only
commit
insert into scheduler_run_details (job_id, start_time, duration, outcome, message, id) values ('SearchAuditListener', '2023-12-17 00:01:49.504', 1, 'S', '', 181438774)
SET autocommit=0
SELECT @@session.transaction_read_only
commit
SET autocommit=0
SELECT @@session.transaction_read_only
commit
SET autocommit=0
SELECT @@session.transaction_read_only
commit
insert into scheduler_run_details (job_id, start_time, duration, outcome, message, id) values ('addonHouskeeperTrigger-v2', '2023-12-17 00:01:49.133', 1, 'S', '', 181438773)
SET autocommit=0
SELECT @@session.transaction_read_only
commit
SET autocommit=0
SELECT @@session.transaction_read_only
commit
SET autocommit=0

[Nic Brough -Adaptavist-]

Welcome to the Atlassian Community!

Off-the-shelf, Confluence has a scheduler configured to run pretty much continuously, so as it needs to log this, writing the updates to the database is normal.

[Yuriy Gurin]

Hello

Yes, it has scheduler and I see it's settings in the confluence's configuration. But the tasks what I see are going minimum each several minutes, but definetly not each second what I'm observing on my instance

EDIT:

I'm observing this behaviour without users logged. Just a running instance, no any user activity