Update ns8-piler to piler 1.4.8

[stephdl]

Update ns8-piler to piler 1.4.8

Proposed solution

Update configs and services to support the latest Manticore and MariaDB versions, as done in PR #34. This will allow us to benefit from new features, improved performance, and better maintainability of the stack. It also ensures compatibility with recent upstream changes and provides a more future-proof foundation for further development.

Alternative solutions

Keep current versions, but this may lead to missing out on improvements and could cause compatibility or support issues in the future.

Additional context

See PR #34 for implementation details.

See also

NethServer/ns8-piler#34

---

thank to @jsuto

[stephdl]

QA

target is ghcr.io/nethserver/piler:1.1.0-dev.3

case 1: direct install

Install a mail server with an account provider and a webmail
install piler from the testing version
configure piler
change the default password of auditor@local and admin@local
with the webmail send email between users
the mails are indexed immediately and visible with the auditor@local user (with the admin@local user you can see the number increasing in the right panel)
test the import-emails : runagent -m piler1 import-emails

case 2: Upgrade

The upgrade will need a lot of care, inspect logs, and check if something is wrong (we add a new manticore container and we upgrade memcached, piler, mariadb)

Install a mail server with an account provider and a webmail
install piler from the stable version
configure piler
change the default password of auditor@local and admin@local
with the webmail send email between users, try to make some data to make a real test (imasync is better a way to make a lod of data)
the mails are NOT indexed immediately and are NOT visible with the auditor@local user (with the admin@local user you cannot see the number increasing in the right panel). You have to wait at least 30 minutes to see them or trigger manually the indexation

podman exec -ti piler-app bash
su - piler
/usr/libexec/piler/indexer.delta.sh
/usr/libexec/piler/indexer.main.sh

verify in the auditor vue the number of indexed emails
Upgrade with the testing version
the upgrade takes care to index the emails again to piler, you will see it in logs
go to the piler website and check with admin@local and auditor@local if you see error in the admin and emails in the auditor page
verify in the auditor vue if you have the same number of indexed emails
with a webmail send emails now , you can see them directly without waiting an indexation, we use the real-time

case3: upgrade a real server

the archive can be lost, with the command runagent -m piler1 import-emails we can send back all the emails found in the server to a new archive
upgrade a real piler instance
or
create a virtual machine with a lot of emails inside (use imapsync and sync thousand of emails)
install and configure piler
import with the script ../bin/import-emails
index the emails

podman exec -ti piler-app bash
su - piler
/usr/libexec/piler/indexer.delta.sh
/usr/libexec/piler/indexer.main.sh

verify in admin and auditor view how many emails you have
upgrade your piler instance

[lucagasparini]

Test case 1 (direct install) : VERIFIED

Test case 2 (upgrade): FAILED
After the upgrade, logging in with auditor does not display any messages.
Messages were displayed in auditor before the upgrade.
New messages after the upgrade are displayed, but old ones are not.