skip to Main Content

Attunity Replicate is a remnant of pre-cloud times

Attunity CDC was first released for IBM mainframe databases in 2004 and eventually evolved into Attunity Replicate. Why dwell on history? Attunity Replicate packages up software from the last decade without any of the technologies taken for granted in modern cloud-first architectures. Thus, Attunity Replicate lacks cloud-first features — such as scale up, scale out, fault tolerance, expandability, elasticity, and others.

Misfit for the cloud

In 2015-2016, Attunity Replicate added support for big data targets like Hadoop and Kafka. However, still missing are common cloud-first building blocks like Kubernetes, Docker, Zookeeper, Elasticsearch, Kibana, Terraform, and others. Also absent is a subscription business model appropriate for the cloud. Under a cloud model, users pay only for resources used without up-front, long term, investments.

Attunity Replicate includes a Binary Reader which parses database redo logs. Such parsers originated when databases did not have redo log mining APIs. Not only are such binary readers obsolete in the cloud age, they are actually a disadvantage to users. This is true because log readers must track database versions. Otherwise, they can break overnight by a database version change. Database vendors do not guarantee backward compatibility in log file formats.

Cloud migration is a specific use case that requires privacy and automation. Attunity Replicate includes no built-in VPN service. Thus, customers moving important corporate data to the cloud need an open source or third-party VPN at additional expense. Users manually copy schema using Microsoft or other tools, start the initial load task, and then engage replication. Plus, users manually check status, triage, and debug after each stage.

A different approach

Griddable takes a completely different approach. Created cloud-first, Griddable provides scale up, scale out, and high availability features out of the box. Due to the importance of on-premise and multi-cloud use cases, Griddable includes a general-purpose IPsec VPN service. Griddable installs on-prem or in the cloud to fit a wide variety of use cases. Regardless of where it is running, its SAAS-based licensing provides flexibility and excellent ROI.

Most importantly, Griddable provides an automated user experience. Simply connect sources and targets to the grid, create a data selection policy, and start flowing data. The user experience is seamless because Griddable copies the source schema, loads initial source data, and engages change capture automatically.

Let’s compare Griddable 5.1 with Attunity Replicate:

Cloud-first

FeatureGriddable 5.1Attunity Replicate
Multi-cloudYes. Can migrate cross-cloud, hybrid or combinations. AWS, VMC, GCE, Azure, Azure StackNo
Cloud-first, elastic, HA and DRYes. With Kubernetes, can add capacityNo. Fixed capacity
Cloud-based monitoringELK based: incoming/apply change rate & throughputTask monitoring, incoming/apply change rate & throughput
Cloud database as sourceYes, Amazon RDS and Aurora, Azure SQL DB as sources No
Security of data through systemYes, support VPNNo. Setup separately
SaaS and manageabilityYes, live patching without installNo. Full reinstall

Data synchronization

FeatureGriddable 5.1Attunity Replicate
Transactional data supportYes, timeline consistentYes, same commit order as source
Policy-basedPolicy language for independent control of both the source and target database. UI-based rules tied to single database
Schema, initial-load, catchupYes. Strong homogeneous schema migration, SSMA, heterogeneous schema migrationSchema migration is poor
TopologiesAny number of sources and targets in M : N topologiesPoint to point only
Live data restructuringRow and column filtering and transformations while data in-flight across the grid.No data transformations
SDK for new database providers Yes, developers can create new source or target connectorsNo
Data sovereignty on sourceYes. Apply policies to either the source or the target, or bothOnly 1 place
Append-only data warehouse Yes, for Hive and othersYes, for Hive-ACID
Error recovery and retry Yes, and configurableYes and configurable
Data comparatorLimited feature. Next release will be full featureNo
Migration dashboardYes, full progress in ELK dashboardLimited, support for task progress bars
DatatypesDon’t support XML. LOBs supported with LogMiner and RAC for all versionsXML via BinaryReader. LOBs need BinaryReader for 12c
Multiple clones, load on sourceNo load on source for copies, with Change History ServerYes, will load source, since log can be purged after it is consumed
In memory transferYesYes
Policy filtering to source to reduce network payloadYes, with GoldenGate. Future release will have split relayNo, replication task reads across WAN
DDL migrationNo, future releaseCreate/drop/rename of table/column

Source and target databases

FeatureGriddable 5.1Attunity Replicate
Source list: All supported Oracle, MySQL, SQLServerAzure SQL DB, Aurora, Oracle RACSybase ASE, DB2, Teradata, Informix, Postgres, Salesforce, SAP, Vertica
Target list: All supported Oracle, MySQL, SQLServerAurora, Azure SQL DB, Hive, HBase, Oracle RAC. Mongo, coming up.Greenplum, Hawq, Actian Vector, Sybase IQ, Azure SQL DW, Sybase ASE, Netezza, MongoDB, Kafka, Teradata Aster, Vertica, Hive ACID, Azure SQL DB

Next step

To see Griddable in action, click the “Demo Now” button for a 10 minute, no-obligation tour.

Back To Top