Main features

This release delivers one of the most substantial architectural upgrades in the product’s history. We have re-engineered core subsystems to push the platform far beyond previous performance ceilings, with a particular focus on indexing, ingestion throughput, and large-scale storage workflows. The result is a faster, smarter, and more resilient Archiver, capable of sustaining heavy workloads with ease while preparing the foundation for future innovations.

New indexing engine

We introduced a redesigned version of our indexing engine that elevates speed, reliability, and operational consistency across the board.

The new engine processes data with reduced latency, supports much faster ingestion rates, and executes queries with higher efficiency—even under sustained load.

Through continuous internal optimizations, smarter memory utilization, and robust failure-resilience strategies, the indexing layer now delivers:

• faster indexing of new items
• more responsive and stable query execution -improved behavior under peak concurrency
• better scaling on large datasets
• lower operational overhead and smoother long-term performance

This is not just an incremental improvement: It’s a foundational leap that enhances nearly every part of the platform relying on search and retrieval.

Fully parallel email ingestion

Until version 25.8, each tenant could process only one email at a time. Even though ingestion was fast, this single-threaded model could still become a bottleneck, especially during first-time imports or heavy bursts of incoming messages.

We have now overhauled this architecture from the ground up. The Archiver can ingest multiple emails simultaneously for the same tenant, unlocking substantial performance gains in real-world workloads.

Key improvements:

• true multi-email parallel ingestion per tenant
• dynamic workload adaptation: the Archiver intelligently detects load and automatically tunes its parallelism to maximize resource usage
• massive improvements in first-time ingestion speed
• better CPU and I/O distribution across the system
• smoother ingestion under sustained or unpredictable spikes

This redesign eliminates the previous serialization point and transforms ingestion throughput into a scalable, high-performance pipeline.

Async volumes and high-speed caching-write Layer

Email archiving involves periodic packaging: every N ingested emails are bundled into an encrypted ZIP file and uploaded to all active data volumes.

In high-ingestion scenarios—especially when uploads are slow, this upload phase could become the limiting factor, forcing the Archiver to wait for storage operations to complete before continuing.

To remove this bottleneck, we introduced a new advanced storage architecture featuring caching-write volumes and async upload volumes.

How it works:

1. Immediate local write Encrypted packages are first written at high speed to a local caching volume, allowing ingestion to continue without interruption.

2. Background asynchronous upload Archiver then handles the distribution of these packages to all active data volumes asynchronously, in the background, without slowing down ingestion.

3. Automatic cleanup Once uploads are successfully completed, files are automatically removed from the caching-write volume.

Benefits:

• upload latency no longer impacts ingestion speed
• dramatically smoother ingestion during heavy workloads
• optimized use of fast local storage
• more predictable throughput, even with slow or remote storage backends
• clear separation between ingestion performance and storage performance

This advanced volume configuration completely eliminates waiting time tied to remote uploads and enables the Archiver to maintain consistently high throughput.

Note: this feature requires appropriate licensing.

Features

• jobs: allow naming jobs to simplify tracking and speed up identification across import, export, and reindex workflows
• jobs page: added pagination, search by job name, and filtering by job type to improve navigation in large job histories
• running job modal: increased its size for better readability, remove the info icon, and allow opening job details by clicking anywhere on the job tile
• new job handling: ensure that newly launched jobs instantly appear as the first entry in the running-job popup for immediate visibility
• job progress percentage: always display two decimal digits to provide clearer insight into the progress of large operations
• volume sections: significantly improved page loading speed, add pagination, enable searching by volume name, and introduce filters by storage type and usage
• volume types: introduced a new dedicated Ionos volume type to support environments using Ionos storage
• reindex metrics: show ZIP download speed, estimated time to download, projected completion date, and extended performance insights
• reindex recovery: allow relaunching failed reindex jobs so that only failed ZIP files are reprocessed
• pst batch import: added a comprehensive panel showing total extracted emails, ingestion rates per hour, imported GB per hour, total archive size, imported size, remaining size, and average email size; include estimated completion date, categorized archive lists (running, pending, completed, failed), and a search panel for locating specific archive files
• pst import: show imported emails per hour and estimated completion date directly within the job view
• export speedup: accelerate exports when emails are stored on local volumes by better leveraging local disk performance
• export metrics: show exported emails per hour, total exported size, average email size, and estimated completion date in the job panel
• volume migration metrics: display total archive size to migrate, migration rate per hour, migrated size, and GB-per-hour throughput
• disk email notification: include disk capacity details in alerts, showing used GB and percentage utilization
• email ingestion page: introduced a new page with historical growth charts for archived emails and storage usage, plus a chart showing recent ingestion rates
• folder structure maintenance: added an automatic scheduled service that keeps mailbox folder structure data clean for faster user navigation
• let’s encrypt: auto-fill the email field with the Libraesva default address when enabling Let’s Encrypt on cloud appliances
• connector error handling: implemented a new system for detecting, parsing, and presenting known errors such as rate limiting
• connector severity tuning: reduced notification severity for temporary conditions like Microsoft 365 rate-limiting
• dns check: introduced a nightly DNS health check to ensure the appliance is resolving correctly
• storage usage report: added a new report tracking storage consumption per tenant over time
• legal hold coverage report: added a report showing how many emails are currently protected by each legal hold rule
• dynamic sizing for database and webserver: introduced automatic resource scaling for both components so that configuration adjusts seamlessly whenever appliance resources are modified
• IMAP and PEC connector validation: refined validation checks to deliver clearer, more actionable feedback when connection issues occur
• volumes validation: updated validation logic to provide more meaningful diagnostics and clearer guidance in case of volume-related problems
• purgeable space update: automatically refresh the displayed purgeable space when a purge job completes, avoiding outdated cached values and ensuring the customer page always reflects accurate data, with periodic updates during the purge process when possible
• search filters: added quick-access filters to easily search emails from today, yesterday, or recent days
• local user aliases: allow creating aliases that point to users managed through the directory service
• job dependencies: added a clear UI status indicating when a job is waiting for another job to complete, with graceful fallbacks using started at and created at timestamps
• NTP configuration: introduced a dedicated page to configure and validate time synchronization settings
• delete protections: added UI errors when attempting to delete emails newer than the latest legal-hold application run
• job termination safety: added a confirmation step before allowing a job to be force-killed
• cluster notifications: added automatic email notification when the cluster setup completes
• volume migration insights: added metrics showing remaining archives, remaining size, and both source and destination volume names
• reboot detection: detect jobs interrupted by a reboot and mark them as failed with a dedicated message
• notifications: made hostnames in email notifications clickable for faster access to the appliance

Security

• local user password: increased security by enforcing a minimum password complexity requirement
• legal hold protection: show a clear warning when attempting to delete emails newer than the most recent legal-hold application job

Improvements

• report email field: improved the input field for entering email addresses so it takes advantage of all available space
• audit log: always display human-readable descriptions instead of raw action entries to make event interpretation clearer
• kill jobs: avoid sending email notifications when jobs are manually terminated from the UI
• login through 365: replaced the Graph API icon with the official Microsoft 365 logo for clearer visual alignment
• snmp section: updated the UI layout to make it easier to understand how to add new SNMP settings
• pst batch import: enhanced validation for unsupported file extensions and provide a clear, specific error message
• disk threshold notifications: reviewed all thresholds, email notification content, and dashboard colors to ensure more consistent and accurate alerting
• external 365 address: automatically exclude external Microsoft 365 addresses from license counting and connector queues to avoid unnecessary processing
• license user count: improved background jobs to increase the accuracy, speed, and resilience of license-usage calculations
• indexing engine: improved cleanup of email content before indexing to prevent unnecessary token indexing and maximize indexing engine performance
• job metadata: clarified the distinction between created at and started at across the job system
• hypervisor console menu: refined the interface for better usability
• export diagnostics: increased logging verbosity to simplify troubleshooting during export operations
• PST import: increased the maximum number of PST files processed concurrently within a batch job to better utilize all available resources
• jobs page performance: switched to asynchronous data loading so the page renders faster, loading job details only upon details panel expansion
• M365 external mailboxes: automatically skip EXT mailboxes when queuing users for the connector
• cluster setup: enhanced the setup flow by guiding users through required firewall settings and validating them proactively
• cluster UI: improved the interface to more clearly present local and remote nodes information
• cluster disk guidance: clarified explanations around cluster-disk usage and detailed the correct configuration procedure
• cluster data model: highlighted that local disk contents are not replicated across nodes and that the passive node’s credentials will be overwritten during cluster setup
• indexing engine: updated the indexing engine page to also display customers without any stored emails
• attachment indexing: improved performance when parsing and indexing Office documents
• version visibility: enhanced the version page to reliably surface newly available minor updates, ensuring clearer upgrade awareness

Bugfix

• purge job UI: fixed a crash that appeared when launching a new purge job from the interface
• purge job stability: fixed a condition where the Archiver could become unresponsive when starting a purge job
• export: fixed an issue that occurred when exporting an excessively large number of files from a single archive

Breaking changes

• legacy search removal: remove the legacy search feature, originally deprecated in v23.6

Update notice

• folder-structure volumes: these volumes are no longer required, as folder-tree management is now fully handled internally by the Archiver. Existing folder-structure volumes are no longer used and can be safely removed manually. In the Libraesva Cloud environment, this cleanup will be performed automatically in the days following the update.

How to update

In Settings > System > General section, you can either update the Archiver immediately or schedule the update for the upcoming night.

Minor updates

–