Smart Update fails during "Analyzing the test site before updating it" at 10%

[Peter van der Kleij]

Server operating system version

AlmaLinux 9.8 (Olive Jaguar)

Plesk version and microupdate number

Plesk Obsidian 18.0.80 Update #6

WP Toolkit 6.11.2
WP Guardian active

WP-Tookit Smart Update, failed after 10% with anylzing.
After some CoPilot testing, if found information in the sqllite database.
These are the results.

Problem
Smart Update consistently fails for the WordPress instance %domain%.
The process reaches:
1: Analyzing the original site OK
2: Creating a test site OK
3: Analyzing the test site before updating it, 10%

and then eventually fails with:
1: Smart Update was unable to correctly clone your website for further analysis.

The same server successfully performs Smart Updates on other WordPress instances.

What has already been verified
The following possible causes have been ruled out:
- Sufficient disk space available
- No quota limitations
- Correct file permissions
- API access restriction for 127.0.0.1 removed
- LiteSpeed Cache disabled
- All In One Security (AIOS) disabled
- WP Toolkit updated to latest available version (6.11.2)
- WP_HOME not defined
- WP_SITEURL not defined

WordPress configuration:
1. wp option get home
2. https://%domain%
3.
4. wp option get siteurl
5. https://%domain%
6.
7. wp option get permalink_structure
8. /%postname%/
9.
10. wp post list --post_status=publish --format=count
11. 12

Clone creation appears successful
During Smart Update a temporary clone directory is created:
1. httpdocs/.wp-toolkit_x

The directory contains a complete WordPress installation:
1. wp-admin
2. wp-content
3. wp-includes
4. ...
5. `

Therefore the cloning stage appears to complete successfully.
SQLite statistics also show:
1. stepCloneSuccess = 1
2. stepClone = 100

WP Toolkit SQLite statistics
Database:
1. /usr/local/psa/var/modules/wp-toolkit/wp-toolkit.sqlite3

Failing instance (%domain%)
Important values recorded:
1. stepInitialCheck = 23
2. stepClone = 100
3. stepCloneSuccess = 1
4.
5. stepPreAnalysis = 0
6. stepHealthCheckPreAnalysis = 0
7.
8. pagesTotalHealthCheck = 0
9.
10. stepUpdate = 0
11. stepPostAnalysis = 0

Working instance on same server
For comparison, another website successfully completes Smart Update and records:
1. stepInitialCheck = 11
2. stepClone = 74
3. stepPreAnalysis = 7
4. stepHealthCheckPreAnalysis = 7
5.
6. pagesTotalHealthCheck = 6
7.
8. stepUpdateSuccess = 1
9. stepPostAnalysisSuccess = 1
10.
11. smartUpdateStatus = success

Observations during clone analysis
During the Smart Update run:
1. curl -I https://%domain%/.wp-toolkit_o/page-sitemap.xml
2.

returns:
1. HTTP/2 200
2. ``

and:
1. curl -s https://%domain%/.wp-toolkit_o/page-sitemap.xml

returns valid sitemap XML.
The clone appears reachable and operational during the analysis phase.
However, WP Toolkit records:
1. pagesTotalHealthCheck = 0

which differs from functioning Smart Update runs on the same server.

Logs
panel.log only reports generic failures referencing:
1. PleskExt\WpToolkit\BackgroundTasksImplementation
2. \Update\Step\SmartUpdateInstance\PreAnalyzing.php

No specific page, URL, shortcode, plugin or PHP error is reported.

Question
Could you please advise:
Which additional logs are available for the Smart Update PreAnalysis / HealthCheck phase?
Why pagesTotalHealthCheck remains 0 for this instance while the clone is successfully created and accessible?
Whether this is a known issue involving Smart Update analysis for specific WordPress installations (Divi / Divi Supreme are in use on this website).

Any guidance to obtain more verbose Smart Update analysis logging would be appreciated.

 

[Sebahat.hadzhi]

@Peter van der Kleij , are there any errors in the browser console when the process gets stuck, please? Also, could you please try temporarily enabling Plesk debug mode and confirm if there are any more verbose entries logged into /var/log/plesk/panel.log? Thanks in advance.