---
title: "Volumes in an OpenSearch cluster"
sidebar_label: "Volumes"
description: "Basic information about the volumes used in an OpenSearch database cluster"
sidebar_position: 2
toc_max_heading_level: 3
---

import Tabs from '@theme/Tabs'
import TabItem from '@theme/TabItem'
import {CustomTable} from '@selectel/docux/components'
import {TabItemLabel} from '@selectel/docux/components'
import Formbricks from '@theme/MDXComponents/Formbricks'

# Volumes in an OpenSearch cluster

In an OpenSearch Managed Database cluster, you can use [local](#local-disks) and [network volumes](#network-volumes). The disk type used depends on the selected [configuration line](/managed-databases/opensearch/configurations.mdx#configuration-lines).

You can select a configuration line with the required disk type and disk size when [creating a Managed Database cluster](/managed-databases/opensearch/create-cluster.mdx). After the cluster is created, you can change the line to another one with the same disk type and increase the disk size — [scale the cluster](/managed-databases/opensearch/resize-cluster.mdx).

Performance and throughput for disks in various configuration lines can be viewed in the [Performance and throughput](#performance-and-bandwidth) table.

## Local disks \{#local-disks}

<CustomTable>
  <table>
    <tbody>
      <tr>
        <th>Configuration lines</th>

        <td>
          <ul>
            <li>Standard;</li><li>CPU;</li><li>Memory;</li><li>Flex</li>
          </ul>
        </td>
      </tr>

      <tr>
        <th>Disk type</th><td>SSD NVMe disk</td>
      </tr>

      <tr>
        <th>Technical details</th><td> The disk subsystem is configured in RAID 10 for the Standard, CPU, Memory, HighFreq, and Flex lines</td>
      </tr>

      <tr>
        <th>Benefits</th><td>Low data access latency</td>
      </tr>

      <tr>
        <th>Usage</th><td>Suitable for tasks sensitive to read and write speed metrics</td>
      </tr>

      <tr>
        <th>Number of disks in DBMS</th><td>1</td>
      </tr>

      <tr>
        <th>Configuration change</th>

        <td>
          <ul>
            <li>in all configuration lines with a local disk, except for Flex, you cannot increase the local disk only; you must change the entire node configuration — [scale the cluster](/managed-databases/opensearch/resize-cluster.mdx);</li>
          </ul>
        </td>
      </tr>

      <tr>
        <th>Disconnecting a disk</th><td>✗</td>
      </tr>

      <tr>
        <th>Connecting to another database</th><td>✗</td>
      </tr>

      <tr>
        <th>Creating another disk, snapshot, or backup</th><td>✗</td>
      </tr>

      <tr>
        <th>Deletion</th><td>Deleted along with all data when the cluster is deleted</td>
      </tr>
    </tbody>
  </table>
</CustomTable>

## Network volumes \{#network-volumes}

<CustomTable>
  <table>
    <tbody>
      <tr>
        <th>Configuration lines</th><td>Flex</td>
      </tr>

      <tr>
        <th>Disk type</th><td>SSD NVMe disk</td>
      </tr>

      <tr>
        <th>Technical details</th><td>Managed by a Software-Defined Storage (SDS) system — Ceph</td>
      </tr>

      <tr>
        <th>Benefits</th>

        <td>
          <ul>
            <li>triple replication of disk volumes;</li><li>network volume size is independent of the number of vCPUs</li>
          </ul>
        </td>
      </tr>

      <tr>
        <th>Usage</th><td>Not recommended for latency-sensitive systems. For example, it is suitable for archive data storage. ( )</td>
      </tr>

      <tr>
        <th>Number of disks in DBMS</th><td>1</td>
      </tr>

      <tr>
        <th>Configuration change</th>

        <td>
          <ul>
            <li>no downtime when changing disk size;</li><li>when changing the number of vCPU and RAM, the downtime depends on the number of nodes in the cluster: in single-node clusters, downtime is equal to the node restart time (one to two minutes), and in multi-node clusters, it is equal to the time required to switch to another node (less than one minute)</li>
          </ul>
        </td>
      </tr>

      <tr>
        <th>Detaching a volume</th><td>✗</td>
      </tr>

      <tr>
        <th>Connecting to another database</th><td>✗</td>
      </tr>

      <tr>
        <th>Creating another volume, snapshot, or backup</th><td>✗</td>
      </tr>

      <tr>
        <th>Deletion</th><td>Deleted along with all data when the cluster is deleted</td>
      </tr>
    </tbody>
  </table>
</CustomTable>

## Performance and throughput \{#performance-and-bandwidth}

<Tabs queryString="performance-and-bandwidth">
  <TabItem value="standard" default>
    <TabItemLabel>
      Standard
    </TabItemLabel>

    <CustomTable>
      <table data-layout-fixed>
        <tbody>
          <tr>
            <th>Throughput</th><td>400 MB/s</td>
          </tr>

          <tr>
            <th>Performance (read)</th><td>25 600 IOPS</td>
          </tr>

          <tr>
            <th>Performance (write)</th><td>12 800 IOPS</td>
          </tr>
        </tbody>
      </table>
    </CustomTable>
  </TabItem>

  <TabItem value="cpu">
    <TabItemLabel>
      CPU
    </TabItemLabel>

    <CustomTable>
      <table data-sticky>
        <thead>
          <tr>
            <th />

            <th>Volumes<br /> up to 256 GB</th><th>Volumes<br /> 512 GB</th>
          </tr>
        </thead>

        <tbody>
          <tr>
            <th>Throughput</th><td>400 MB/s</td><td>500 MB/s</td>
          </tr>

          <tr>
            <th>Performance (read)</th><td>25 600 IOPS</td><td>40 000 IOPS</td>
          </tr>

          <tr>
            <th>Performance (write)</th><td>12 800 IOPS</td><td>20 000 IOPS</td>
          </tr>
        </tbody>
      </table>
    </CustomTable>
  </TabItem>

  <TabItem value="memory">
    <TabItemLabel>
      Memory
    </TabItemLabel>

    <CustomTable>
      <table data-sticky>
        <thead>
          <tr>
            <th />

            <th>Volumes<br /> up to 256 GB</th><th>Volumes<br /> 512 GB</th>
          </tr>
        </thead>

        <tbody>
          <tr>
            <th>Throughput</th><td>400 MB/s</td><td>500 MB/s</td>
          </tr>

          <tr>
            <th>Performance (read)</th><td>25 600 IOPS</td><td>40 000 IOPS</td>
          </tr>

          <tr>
            <th>Performance (write)</th><td>12 800 IOPS</td><td>20 000 IOPS</td>
          </tr>
        </tbody>
      </table>
    </CustomTable>
  </TabItem>

  <TabItem value="flex">
    <TabItemLabel>
      Flex
    </TabItemLabel>

    <CustomTable>
      <table data-sticky>
        <thead>
          <tr>
            <th />

            <th>Local disk</th><th>Network volume (except ru-1, ru-2,<br />ru-3, ru-6)</th><th>Network volume (ru-1, ru-2, ru-3)</th><th>Network volume (ru-6)</th>
          </tr>
        </thead>

        <tbody>
          <tr>
            <th>Throughput</th><td>300 MB/s</td><td>500 MB/s</td><td>500 MB/s</td><td>1000 MB/s</td>
          </tr>

          <tr>
            <th>Performance (read)</th><td>12 800 IOPS</td><td>25 000 IOPS</td><td>12 800 IOPS</td><td>75 000 IOPS</td>
          </tr>

          <tr>
            <th>Performance (write)</th><td>6 400 IOPS</td><td>15 000 IOPS</td><td>6 400 IOPS</td><td>75 000 IOPS</td>
          </tr>
        </tbody>
      </table>
    </CustomTable>
  </TabItem>
</Tabs>

<Formbricks />