Lindorm Tunnel Service (LTS) moves data into, out of, and between Lindorm databases. LTS supports full and incremental data migration, real-time change tracking, data lake archiving, cross-cluster synchronization, and backup and restore.
Core capabilities
|
Capability |
Description |
|
Cloud-native and distributed |
Runs on Elastic Compute Service (ECS) instances with horizontal scaling. Add nodes on demand to match your workload without pre-provisioning resources. |
|
Single-step configuration |
Select a source, a destination, and the columns to synchronize. LTS automatically replicates the schema and runs full and incremental synchronization — no manual schema mapping or phased setup required. |
|
Built-in safety checks |
Before a task starts, LTS checks network connectivity and security automatically. During the task, it monitors synchronization latency and destination storage usage in real time, applying throttling and alerts as needed. After the task completes, LTS verifies data accuracy. |
|
Optimized for HBase and Cassandra |
File-level data processing delivers up to 10x the throughput of API-layer migration, with CPU, cache, memory, and network I/O tuned to lower transfer costs. |
Supported data paths
|
Source |
Destination |
Use cases |
References |
|
HBase |
LindormTable |
Data migration between clusters, cluster upgrades, online and offline workload decoupling, primary/secondary disaster recovery, and active geo-redundancy |
|
|
MaxCompute / Hive |
LindormTable |
Offline query acceleration, pushing data warehouse details and metrics to Lindorm for online queries |
Contact technical support |
Real-time change tracking
Lindorm Stream provides real-time incremental data change subscriptions in Lindorm. For details, see Lindorm Stream.
Deprecated features
The following features are no longer available for LTS instances purchased after the dates shown. If your instance was purchased before the cutoff date, you can still use the feature.
|
Feature |
Cutoff date |
Alternative |
|
ApsaraDB RDS to LindormTable (full and incremental import) |
March 10, 2023 |
Use Data Management (DMS) to synchronize data. See Import full data and incremental data from ApsaraDB RDS for existing instances. |
|
LindormTable to MaxCompute (full and incremental export) |
June 16, 2023 |
See Export full data to MaxCompute and Archive incremental data to MaxCompute for existing instances |
|
Simple Log Service (SLS) LogHub subscription (incremental import) |
June 16, 2023 |
See Import incremental data from SLS for existing instances |
Use cases
Cluster migration
Migrate data from HBase to Lindorm, switch cluster networks — for example, from the classic network to a Virtual Private Cloud (VPC) — move data across regions, or decouple workloads between clusters, all without service interruption.
How it works:
LTS reads historical data directly from the Hadoop Distributed File System (HDFS) of the source cluster, bypassing the HBase or Lindorm service layer. Production workloads remain unaffected.
Within the same task, LTS synchronizes real-time incremental data so the destination stays current throughout the migration.
After migration completes, LTS verifies data accuracy and confirms consistent partitions through automatic schema synchronization.
Performance:
Each node migrates data at up to 100 MB/s. Add nodes to scale linearly for terabyte- or petabyte-scale data volumes.
File-level processing cuts data usage by more than 50% compared to API-layer migration.
Built-in retry logic, real-time progress monitoring, and failure alerts keep the process stable.
Online and offline workload decoupling
Synchronize online business data in real time to Hadoop Distributed File System (HDFS) or Object Storage Service (OSS), then analyze it with Spark or MapReduce without affecting online query performance.
Primary/secondary disaster recovery
Set up two-way data synchronization between an active cluster and a standby cluster. If the active cluster fails, switch to the standby cluster to minimize downtime. After recovery, use LTS to synchronize incremental data from the standby cluster back to the active cluster.
Historical data archiving from ApsaraDB RDS
As transaction data grows in ApsaraDB RDS, query performance degrades and storage costs rise. LTS synchronizes data from ApsaraDB RDS to LindormTable in real time, separating hot data from cold data to address both problems.
LindormTable handles the archived data with automatic horizontal scaling, high-concurrency queries, multi-dimensional indexing, and lightweight analytics. Lindorm Stream tracks data changes in sequence. LTS also synchronizes data from LindormTable to other analytics systems for further analysis.
ApsaraDB RDS to Lindorm synchronization is only available for LTS instances purchased before March 10, 2023. For newer instances, use Data Management (DMS) instead.