Features
What AegisGrid can do
More than backup and recovery: a backup system for servers, virtual machines and Microsoft 365 – deduplicated, encrypted, to several destinations at once, processed further on the backup server, restored as a test and logged with proof. From a USB disk to the data centre.
Back up
- Files, volumes, partitions, whole disks and images
- Content-defined deduplication across all backups of a store
- Every version complete and restorable on its own, without backup chains
- Free space (NTFS, ext2/3/4) and zero areas are not stored
- Compression with zstd or lz4
- Parallel processing for high throughput
Consistency and platforms
- Windows with VSS shadow copies
- Linux with LVM snapshots
- Pre/post hooks for databases and applications
- Hyper-V: whole VMs, application-consistent, with change tracking
- Proxmox VE (LVM, LVM-thin, ZFS, Ceph RBD, qcow2) and Xen: whole VMs with their configuration
- Proxmox VE with fast follow-up backups (CBT): after the first backup, only what changed in the VM is read
- Xen with memory if wanted: the restored VM carries on where it stood at the backup
- Microsoft 365: Exchange Online, OneDrive, SharePoint
Back up files and folders
- Choose folders and files in a tree, with your own and sensible default exclusions
- Permissions and owners included: NTFS permissions, POSIX ACLs, extended attributes
- Windows with VSS and the change journal: only what has changed is read
- Linux with an LVM snapshot or live, detecting files that change while they are read
- Deduplicated together with the disk and VM backups of the same store
- Warning when unusually many files have changed – older versions are then not cleaned up
Get single files back
- Search by name across all versions, optionally for a period of time
- Every version of a file and the comparison of two versions
- To the original location, to another folder or as a ZIP archive
- Open a version as a read-only folder, on Windows and Linux
- Tick files and folders in the management console
- Single files from disk and VM backups as well, on Windows and Linux, without restoring the disk first
Security
- AES-256-GCM or ChaCha20-Poly1305 per store
- Encryption before transfer, the key stays with the customer
- TLS with a pinned server certificate
- API keys bound to directories
- Backup server with separated privileges
- Ed25519-signed updates and components
Protection against loss and tampering
- WORM storage with a retention lock
- Simply back up straight to a USB disk or a NAS – without a backup server
- Up to eight destinations in one run: local disk, NAS, your own backup server at a second site, the AegisGrid cloud storage or S3-compatible object storage
- On S3 always encrypted, immutable through S3 Object Lock if wanted
- Mirroring of whole stores to a second location
- Export of single backups as an archive for tape or USB disk
- Retention per destination, also by generations (days, weeks, months, years)
- Extend the retention of single backups on purpose, for an audit, for example
Check and restore
- Check of all stored objects (scrub)
- Comparison of a backup with its snapshot (Linux)
- Repair from a second copy
- Partial restore of damaged backups
- VMs as a new VM or disk by disk, straight from the server
- Reset Proxmox and Xen VMs in place to an earlier version; the previous disks are kept until the restore has succeeded
- Write back only changed blocks – much faster after a small change
- Host set up anew: VMs come back with their old VMID or name and their MAC addresses
- Restore wizard in the management console: a drive or a VM back step by step, with the destination checked first
- Restore in the background
- Check restore on the backup server: VMs start as a test in an isolated sandbox
- Restore from your own backup servers and destinations even after the licence has expired
- Virus scan of the backups in the check restore, naming the last backup without the finding
Plan and operate
- Job scheduler as a Linux and Windows service, parallel jobs
- Schedules with public holidays per region, time windows and catching up on missed runs
- Dependencies between jobs, schedule editor and the schedule in words
- Results by email; changes through the API with the changed lines
- Failed jobs are retried automatically, with a warning when a job has not succeeded for too long
- Entries in the Windows event log for monitoring
- Management console for Windows; release to the consoles of the other computers in the company
- Web interface for job scheduler and backup server: operation in the browser, without Windows and without additional software – also for pure Linux, Proxmox and Xen environments
- Programs, consoles, messages, e-mails and reports in English, German, Spanish, French and Croatian; new translations arrive with the updates
- HTTPS API to create, start, pause and monitor
- Pause running backups and continue at the same point, bandwidth limit
- Resumption after standby, power-off or a lost connection
- Jobs on a schedule or by hand only
- Updates hold running backups and continue them afterwards
- Wizard for new jobs: source, destinations and schedule in three steps
- After installation, the setup leads straight to the first backup
- F1 opens the online manual at the matching section
- Notice of virtual machines that no job backs up yet
Backup server that processes backups further
- Jobs on incoming backups, on a schedule or one after the other in flows
- Check, mirror to a second location, copy to S3, export, clean up, your own scripts
- Check restore: check file systems, read every file, report signs of ransomware, start VMs; check report with traffic lights and screenshots
- Messages per customer to the customer, the partner or both
- Several customers cleanly separated, quotas per customer
- Target console for Windows with history and a graphical editor for flows and check chains
- Operation in the browser as well, through the built-in web interface
- Limits for CPU time, memory and disk access per job; usage measured per customer
Evidence for NIS2, critical infrastructure and BSI
- A log of everything that happens to the data
- Chained with SHA-256, every day signed with Ed25519
- Copy to a SIEM via syslog (RFC 5424, also over TLS) or to a second backup server
- Signed evidence report as a PDF every month, also per customer
- Export as CSV or JSON for auditors
Interruptions cost no work
If a backup is interrupted – because a notebook is closed, the battery runs flat, a computer shuts down or the connection drops – AegisGrid carries on afterwards where it stopped. After a lost connection it continues the same backup; after a power-off it picks it up again by itself at the next start.
Only what is still missing is transferred, and a backup only counts as finished once it is complete. Resumption is always switched on.
Even a deliberately held backup carries on at the same point: virtual machines keep their checkpoint or snapshots, finished disks and files are not read again, and the backup keeps the moment of its first start.
Especially for large amounts of data over slow lines: even a first backup that does not run through in one go reaches its destination safely over several stages – without starting from scratch each time.
Your own operation or Managed Backup
You can run AegisGrid yourself – on your own servers, with your own backup server and your own destinations. Or you leave the operation to us: with Managed Backup we do the installation, the updates of the backup servers and backup agents and regular restore tests, wherever your backups are kept. As the destination for the off-site copy, the cloud storage in certified data centres in Germany is also open to you.
Does AegisGrid fit your environment? On Proxmox VE, AegisGrid backs up virtual machines on LVM, LVM-thin, ZFS, Ceph RBD and qcow2 files; snapshots of single Linux servers are taken through LVM. In a pilot project we check beforehand whether all your systems are covered.
See for yourself
We show you AegisGrid on your own systems.