What Is an H3C CAS/UIS Backup Solution?
An H3C CAS/UIS backup solution is a data protection product that captures and restores virtual machines running on H3C CAS (Cloud Automation System) and H3C UIS (Unified Infrastructure System). Both are H3C's in-house virtualization platforms, widely deployed in government, finance, education, and enterprise data centers.
H3C CAS is built on two components: CVK (Cloud Virtualization Kernel), a KVM-based hypervisor that virtualizes compute, and CVM (Cloud Virtualization Manager), the management layer that handles virtual compute, network, storage, high availability, DRS, and built-in backup and recovery. H3C UIS extends CAS into a hyper-converged appliance, adding the ONEStor distributed storage layer so that compute and storage live on the same nodes.
A genuine H3C CAS/UIS backup solution works at the hypervisor level. Rather than installing an agent inside every guest, it reads VM images through the platform's APIs and stores restorable copies on separate storage. This matters because the whole VM — OS, applications, and data — is packaged as files that can be captured and recovered as one unit.
Why You Need to Back Up H3C CAS/UIS
Virtualization consolidates hundreds of workloads onto a few hosts, which raises the stakes when something goes wrong. A single failed host, storage fault, or ransomware event can now take down many VMs at once.
Ransomware and malware. An infected VM can encrypt shared storage, and in a hyper-converged UIS cluster, the blast radius extends across all nodes.
Human error. A wrong operation in CVM or UIS management can delete a VM or its disks with no recycle bin.
Hardware and storage failure. A failed CVK host or a ONEStor reconstruction event can threaten many workloads simultaneously.
Compliance and retention. Industries such as finance and government (a core H3C market) require provable, long-term, restorable retention.
Built-in backup is not enough. H3C’s native backup covers basic full/incremental/differential jobs but lacks the enterprise features — granular recovery, offsite copies, automated verification.
Critically, H3C’s own documentation warns that VMs using raw (high-speed) disks or block devices do not support online backup, incremental backup, or differential backup. Relying on the built-in tool alone leaves exactly the most demanding workloads unprotected.
Key Backup Challenges in H3C CAS/UIS
| Challenge | Why It Matters |
|---|
| Raw/high-speed disk limits | H3C CAS does not support online, incremental, or differential backup for raw‑file or block‑device VMs — these need a workaround or a third‑party tool. |
| Application consistency | Without quiescing, databases and transactional apps can be captured mid‑write, producing an unusable restore point. |
| No native CBT on CAS | Unlike VMware’s Changed Block Tracking, H3C CAS has no equivalent, so a third‑party tool must provide its own changed‑block mechanism (such as Vinchin’s SpeedKit) for fast incrementals. |
| ONEStor distributed storage | In UIS clusters, backup must coexist with the SDS layer without overloading nodes during reconstruction or peak I/O. |
| Backup window and host load | H3C documentation advises against running backups during peak hours because they consume CVK host resources. |
| Limited built‑in recovery | Native CAS backup provides basic full‑VM restore but lacks granular file‑level recovery and instant recovery. |
| Hybrid and multi‑platform estates | Many H3C sites also run VMware, Proxmox, or physical servers, so a tool that only protects H3C creates management silos. |
3 Common Methods to Protect H3C CAS/UIS
There are three practical ways to protect an H3C CAS/UIS environment, with very different trade-offs.
Method 1: Built-in H3C Backup
H3C CAS and UIS ship with a native backup module accessible from CVM. It supports full, incremental, and differential backups to a local directory or a remote server, with scheduled backup policies. This is convenient for small, non-critical environments.
However, the built-in tool has clear limits: no online/incremental/differential backup for raw disks, no granular file recovery, no instant recovery, and no offsite or cloud archiving. It is best treated as a basic first step, not a complete data protection strategy.
Method 2: Storage-Level Snapshots and Replication
On UIS, the ONEStor layer provides snapshots, clones, replication, and erasure-coding redundancy, plus optional CDP (continuous data protection) and SRM-based asynchronous replication for disaster recovery. These protect against hardware and storage faults.
Snapshots and replication, however, are not backups: they live on the same system and are not application-aware. They are excellent for fast rollback and DR, but they do not replace a separate, restorable backup copy kept off the production platform.
Method 3: Third-Party Agentless Backup
The recommended approach for production is a purpose-built, H3C-compatible backup product that performs agentless, image-level backup of CVK-hosted VMs. These tools add what the built-in backup lacks: application-consistent quiescing, changed-block incrementals, granular and instant recovery, offsite copies, and automated verification — all from a single console.
Because H3C positions itself as an open platform, certified third-party tools integrate through its northbound APIs. This is the only method that meets the enterprise requirements in Section 5.
| Method | Agentless | Granular Recovery | Offsite/Cloud | Instant Recovery | Suitable for |
|---|
| Built‑in H3C backup | Yes | No | No | No | Small, non‑critical environments |
| ONEStor snapshot/replication | Yes | No | Partial (DR) | No | Rollback and disaster recovery |
| Third‑party agentless backup | Yes | Yes | Yes | Yes | Production data protection |
Best Practices for H3C CAS/UIS Backup
Use agentless, image-level backup. Protect the whole VM at the hypervisor level to avoid per-guest agent overhead and compatibility issues.
Plan backups outside peak hours. H3C documentation explicitly warns against peak-time backups, so schedule jobs for low-activity windows and use rate limiting.
Handle raw-disk workloads explicitly. Because raw disks cannot be incrementally backed up by the native tool, use a third-party tool that supports them or plan for full backups with a dedicated schedule.
Enable application-consistent quiescing. Coordinate with guest tools to ensure databases and transactional apps restore cleanly.
Adopt forever-incremental with changed-block detection. Minimize storage and time by backing up only changed blocks after the first full backup.
Follow the 3-2-1 rule. Keep at least three copies on two media types, with one offsite — especially important for H3C’s government and finance customers with compliance needs.
Verify recoverability. Automatically boot backups in an isolated lab on a schedule, and run regular restore drills.
Centralize multi-platform protection. Choose a tool that also covers VMware, Proxmox, or physical servers to avoid silos in hybrid H3C estates.
What to Look For in an H3C Backup Tool
When evaluating a third-party H3C CAS/UIS backup solution, use this short checklist. It should support:
Agentless image-level backup through CVK
Provide its own changed-block detection (since CAS has no native CBT)
Handle raw-disk VMs that the built-in tool cannot increment
Offer granular file recovery and instant VM recovery
Keep offsite or cloud copies for 3-2-1 compliance and protect hybrid estates that also include VMware, Proxmox, or physical servers.
Certification - such as H3Cloud Ready - is a strong signal that the vendor has been tested for seamless integration with the platform.
How Vinchin Protects H3C CAS/UIS
Vinchin Backup & Recovery is a purpose-built, agentless backup solution that is certified for H3C — holding H3Cloud Ready certification and partnering with H3C on a joint whitepaper for centralized backup and seamless recovery. It protects standalone and clustered H3C CAS/UIS hosts from a single web console.
Agentless, API-based backup. Vinchin captures H3C CAS/UIS VMs through the hypervisor with no in-guest agents, reducing overhead and maintenance costs.
SpeedKit changed-block detection. Because H3C CAS lacks native CBT, Vinchin’s SpeedKit identifies changed data blocks from snapshots, enabling fast forever-incremental backups.
Instant VM Recovery. Vinchin can roll a critical H3C VM back to a working state in about 15 seconds by mounting backup storage over NFS, dramatically shortening RTO.
Storage efficiency. Deduplication (including zero-block detection), compression, and BitDetector — which excludes swap files, partition gaps, and unpartitioned space — minimize backup footprint.
Granular and cross-platform recovery. Restore individual files without a full VM, or migrate workloads across platforms (V2V, P2V, C2V, V2C), such as from H3C CAS/UIS to VMware.
Offsite copy and cloud archive. Copy backups to a second site or archive to AWS S3, Azure Blob, Alibaba Cloud, or Tencent Cloud for 3-2-1 resilience and long-term retention.
Security and verification. AES-256 / SM4 encryption at rest, RSA / SM2 in transit, anti-ransomware storage protection, and automatic recovery verification keep backups tamper-resistant and provably restorable.
Key Takeaways:
H3C CAS/UIS backup must handle raw-disk limits and the lack of native CBT.
The built-in H3C backup is a start, not a strategy.
Agentless, changed-block-aware backup with offsite copies and verified recovery is the right production approach.
Vinchin delivers H3Cloud-certified, agentless protection with SpeedKit, 15-second instant recovery, BitDetector, and cross-platform migration.
FAQ
Q1: What is the difference between H3C CAS and H3C UIS?
A1: H3C CAS is the virtualization platform (CVK hypervisor + CVM manager). H3C UIS is the hyper-converged appliance that bundles CAS with the ONEStor distributed storage layer on the same nodes.
Q2: Can I back up H3C CAS VMs without installing agents?
A2: Yes. Agentless backup reads VM images through the platform’s APIs at the hypervisor level, so nothing needs to be installed inside each guest.
Q3: Does H3C CAS support incremental backup for all disks?
A3: No. H3C’s documentation states that VMs using raw (high-speed) files or block devices do not support online, incremental, or differential backup via the built-in tool.
Q4: Does H3C CAS have Changed Block Tracking (CBT)?
A4: Not natively. A third-party tool must supply its own changed-block mechanism — Vinchin uses SpeedKit for this purpose — to enable fast incremental backups.
Q5: Is a ONEStor snapshot the same as a backup?
A5: No. Snapshots and replication protect against hardware faults and enable rollback, but they are not application-aware and are not a separate, restorable copy stored off the platform.
Q6: How fast can Vinchin recover an H3C CAS/UIS VM?
A6: Instant VM Recovery can restart a VM in about 15 seconds in typical scenarios by mounting backup data over NFS, depending on storage, network, and VM size.
Q7: Can Vinchin migrate H3C CAS/UIS VMs to VMware?
A7: Yes. Vinchin supports cross-platform V2V recovery, re-converting VM images between H3C CAS/UIS and other hypervisors such as VMware or Proxmox.
Q8: How do I verify my H3C backups are restorable?
A8: Use automatic verification that boots backups in an isolated lab on a schedule, and run regular restore drills - a core requirement for government and finance compliance.
Conclusion
H3C CAS and H3C UIS underpin a large share of China’s virtualized data centers, and protecting them requires a strategy — not just a checkbox. The built-in backup is a useful starting point, but its raw-disk limits, lack of changed-block tracking, and absence of granular or instant recovery leave production workloads exposed.
A robust H3C CAS/UIS backup solution should be agentless and changed-block-aware, support application-consistent quiescing, deliver granular and instant recovery, maintain offsite and cloud copies, and prove recoverability through automatic verification. Vinchin Backup & Recovery, with H3Cloud Ready certification and a joint whitepaper with H3C, meets these requirements in a single, unified, cost-effective console — making it a natural choice for organizations that depend on H3C infrastructure.
