Cloudera Docs

OpDB system requirements

Before you start using the Accumulo parcel, review the prerequisites and system requirements to ensure all requirements are fulfilled.

  • Accumulo depends on HDFS and ZooKeeper libraries and configuration information. Therefore, HDFS and ZooKeeper are mandatory services for OpDB.
  • Accumulo TabletServers must be collocated with DataNodes.
  • Optionally, you can use Accumulo with YARN (MapReduce 2).
Table 1. OpDB system requirements for Accumulo-on-cdp 2.1.2 parcel
Supported versions Note
CDP

CDP Private Cloud Base from 7.1.9

CDP Public Cloud is not supported.

CDP Private Cloud Data Services are not supported.
Operating System

RHEL 7, RHEL 8, and RHEL 9

Ubuntu 20.04 Focal Fossa

SUSE Linux Enterprise Server 12 or 15

Supported on the operating system with the same minor versions that are supported for the CDP installation with which OpDB is installed.
JDK

Java 11 and Java 17

Supported for the CDP install with which OpDB is installed.
Table 2. OpDB system requirements for Accumulo-on-cdp 1.1.0 parcel
Supported versions Note
CDP

CDP Private Cloud Base 7.1.7 SP1 and higher

CDP Public Cloud is not supported.

CDP Private Cloud Experiences are not supported.
Operating System

RHEL 7 / RHEL 8

Ubuntu 18.04 Bionic Beaver

Ubuntu 20.04 Focal Fossa

SUSE Linux Enterprise Server 12

Supported on the operating system with the same minor versions that are supported for the CDP installation with which OpDB is installed.
JDK

Java 8 and Java 11

Supported for the CDP install with which OpDB is installed.
Table 3. OpDB system requirements for Accumulo-on-cdp 1.10.3_7.1.7.2000.0 parcel
Supported versions Note
CDP

CDP Private Cloud Base 7.1.7 SP2 and higher

CDP Public Cloud is not supported.

CDP Private Cloud Experiences are not supported.
Operating System

RHEL 7 / RHEL 8

Ubuntu 18.04 Bionic Beaver

Ubuntu 20.04 Focal Fossa

SUSE Linux Enterprise Server 12

Supported on the operating system with the same minor versions that are supported for the CDP installation with which OpDB is installed.
JDK

Java 8 and Java 11

Supported for the CDP install with which OpDB is installed.
Table 4. OpDB system requirements for Accumulo-on-cdp 1.10.3_7.1.7.2000.1 parcel
Supported versions Note
CDP

CDP Private Cloud Base 7.1.7 SP2 and higher

CDP Public Cloud is not supported.

CDP Private Cloud Experiences are not supported.
Operating System

RHEL 7 / RHEL 8 / RHEL 9

Ubuntu 18.04 Bionic Beaver

Ubuntu 20.04 Focal Fossa

SUSE Linux Enterprise Server 12 / SUSE Linux Enterprise Server 15

Supported on the operating system with the same minor versions that are supported for the CDP installation with which OpDB is installed.
JDK

Java 8 and Java 11

Supported for the CDP install with which OpDB is installed.
Table 5. OpDB system requirements for Accumulo-on-cdp 1.1.1 parcel
Supported versions Note
CDP

CDP Private Cloud Base 7.1.7 SP1 and higher

CDP Public Cloud is not supported.

CDP Private Cloud Experiences are not supported.
Operating System

RHEL 7 / RHEL 8 / RHEL 9

Ubuntu 18.04 Bionic Beaver

Ubuntu 20.04 Focal Fossa

SUSE Linux Enterprise Server 12 / SUSE Linux Enterprise Server 15

Supported on the operating system with the same minor versions that are supported for the CDP installation with which OpDB is installed.
JDK

Java 8 and Java 11

Supported for the CDP install with which OpDB is installed.
Table 6. Hardware requirements for Accumulo parcels
Component Java Heap CPU Disk
Master

Minimum: 1 GB

  • 100-1000 tablets: 4 GB
  • 10,000 or more tablets with 200 or more Tablet Servers: 8 GB
  • 10,000 or more tablets with 300 or more Tablet Servers: 12 GB

Set this value using the Tablet Server Max Heapsize Accumulo configuration property.

See Flume Memory Consumption.

 2 Cores. Add more for large clusters or bulk load. 1 disk for local logs
Tablet Server
  • Minimum: 512 MB Java Heap
  • Recommended: 8 GB
  • Medium scale production (200 or more tablets per Tablet Server): 16 GB

Set this value using the Tablet Server Max Heapsize Accumulo configuration property.

 Minimum 4 dedicated cores. You can add more cores for larger clusters, when using replication, or for bulk loads.
  • 4 or more spindles for each HDFS DataNode
  • 1 disk for local logs (this disk can be shared with the operating system and/or other Hadoop logs
Tracer

1 - 2 GB, depending on cluster workloads.

Set this value using the Tracer Max Heapsize Accumulo configuration property.

 2 or more dedicated cores, depending on cluster size and workloads 1 disk for local logs, which can be shared with the operating system and/or other Hadoop logs
GC Role

1 - 2 GB

Set this value using the Garbage Collector Max Heapsize Accumulo configuration property.

 2 or more dedicated cores, depending on cluster size 1 disk for local logs, which can be shared with the operating system and/or other Hadoop logs
Monitor Role

1 - 2 GB

Set this value using the Monitor Max Heapsize Accumulo configuration property.

 2 or more dedicated cores, depending on cluster size and workloads 1 disk for local logs, which can be shared with the operating system and/or other Hadoop logs