Hello Team
While doing multiple deletions of book’s, pressbooks throws an error “The source books taxonomies could not be read” for subsequent cloning activities.
Link: Business Operations Analytics – Open Textbook
This is observed exclusively in Pressbooks Bedrock Installations.
Using chatgpt, we could figure that some wp_1_* tables (database) were getting deleted and our diagnosis of the database reflects the same. It did give a short script to recover, but this issue keeps happening whenever an user deletes a book, making the recovery pointless, as we need to run the recovery script whenever a user deletes.
Recovery script:
root@e2e-32-111:~# ls
hosting-tools mysql.txt
root@e2e-32-111:~# client_loop: send disconnect: Broken pipe
ugen@Ugendreshwars-MacBook-Air ~ % cd pres
cd: no such file or directory: pres
ugen@Ugendreshwars-MacBook-Air ~ % cd pressbooks
ugen@Ugendreshwars-MacBook-Air pressbooks % ls
download post.sh
install_pressbooks_bedrock.sh restore.sh
ugen@Ugendreshwars-MacBook-Air pressbooks % vi restore.sh
Restart PHP-FPM & Apache and flush redis + WP cache
systemctl restart “php8.3-fpm” || true
systemctl restart apache2 || true
redis-cli FLUSHALL || true
sudo -u www-data wp cache flush --path=/var/www/pressbooksoss-bedrock/web/wp || true
echo “Services restarted and caches flushed.”
show created tables
mysql -NBe “SHOW TABLES LIKE ‘wp_1_%’;” pressbooks
try updating option in site 1 via wp-cli
sudo -u www-data wp option update blogname “Pressbooks Main Site (repaired)” --url=“https://pressbooks.qbnox.com” --path=/var/www/pressbooksoss-bedrock/web/wp || true
check current blog id and site object again
APP_DIR=/var/www/pressbooksoss-bedrock
WP_PATH=“${APP_DIR}/web/wp”
sudo -u www-data wp eval ‘echo “get_current_blog_id(): “.get_current_blog_id().”\n”; print_r(get_current_site());’ --path=“$WP_PATH” | sed -n ‘1,120p’
Hi @Ugendreshwar_Kudupud i’ve read the description but don’t understand what’s happening yet. Book deletion and book cloning are completely separate activities, so this opening sentence confuses me: “While doing multiple deletions of book’s, pressbooks throws an error “The source books taxonomies could not be read” for subsequent cloning activities.”
Where are you experiencing the error? While deleting books? After deleting books on an admin page? While trying to clone books?
Let’s leave the ChatGPT suggestions aside until we clearly understand the error you’re experiencing. Can you
- provide us a simple descriptive summary of the precise steps you’re taking that lead to this error message (along with screenshots or a brief recording) so that we or others can try to replicate the problem?
- provide us with the contents of networks’ diagnostics page, which you can access at Sign in to your account
That information will help other forum users reproduce and advise you on how this problem might be able to be solved.
The following are the steps.
Step 1: Redis cache has been enabled.
Step 2: PB Users clone a couple of books from the public PressBooks sites.
Step 3: One User from their login deletes a cloned book.
Step 4: Same User clones a book from this specific URL Business Operations Analytics – Open Textbook
Step 5: Observes the error “The source book taxonomies could not be read“
Step 6: Me as admin user, clear’s the Redis Cache and the User tries to clone the book once again from the location. Business Operations Analytics – Open Textbook
Step 7: The error is gone and book clones for the user. The same error is reproducible even with admin user also.
The above steps are reproducible multiple times only when Redis cache is enabled.
Observations from the syslog are that WP is looking to access some blog-related tables with a -1 index instead of the blog id (like 1 2 3 etc) index during the taxonomies look-up/set-up.