Docs Menu

Fix IOPS Issues

On this page

  • Alert Conditions
  • Common Triggers
  • Fix the Immediate Problem
  • Implement a Long-Term Solution
  • Monitor Your Progress

Disk I/O % Utilization alerts indicate that percentage of time during which requests are being issued reaches a specified threshold. This threshold is specified when the alert is created.

Note

The utilization measurements for the following alerts include requests from all processes, not just MongoDB processes.

You can configure the following alert conditions in the project-level alert settings page to trigger alerts.

Disk I/O % utilization on Data Partition occurs if the percentage of time during which requests are being issued to any partition that contains the MongoDB collection data meets or exceeds the threshold.

Disk I/O % utilization on Index Partition occurs if the percentage of time during which requests are being issued to any partition that contains the MongoDB index data meets or exceeds the threshold.

Disk I/O % utilization on Journal Partition occurs if the percentage of time during which requests are being issued to the partition that contains the MongoDB journal meets or exceeds the threshold.

A few common events may lead to high Disk I/O % Utilization and trigger these alerts:

Consider a few possible actions to help resolve Disk I/O % Utilization alerts:

  • Optimize your queries.
  • Use the Atlas Performance Advisor to view slow queries and suggested indexes.
  • Review Indexing Strategies for possible further indexing improvements.

    Note

    You may need to temporarily increase your cluster IOPS to create new indexes. To change a cluster's IOPS, go to the Cluster Configuration page and:

    Cloud Provider
    Tier

    Possible Actions

    AWS

    M10, M20

    AWS

    M30 or larger

    Google Cloud

    M10 or larger
    Azure
    M10 or larger
  • Analyze Query Performance to review how your queries are using your indexes.
  • Increase hardware resources, such as instance size and IOPS, in the Cluster Configuration Page.

Atlas clusters deployed to AWS that meet the following criteria may use burst credits when exceeding configured IOPS thresholds during periods of high disk I/O:

  • Clusters where the Provision IOPS setting is disabled.
  • Clusters configured with less than 1 TB of storage capacity.

    Note

    By default, all clusters smaller than the M140 cluster tier are configured with less than 1 TB of storage capacity.

AWS uses burst credits to temporarily increase IOPS beyond the configured threshold during periods of high disk I/O activity. These clusters refresh spent credits during periods of low disk I/0 activity. When burst credits are exhausted, AWS limits IOPS for these clusters to the configured threshold. This may result in decreased disk I/O performance on these clusters and a high IOWait curve in the Normalized System CPU metric.

Atlas clusters deployed to Azure may also use credit-based bursting, but the disk will burst only if it has burst credits accumulated in its credit bucket. Azure also offers an on-demand bursting model, where the disk bursts whenever its needs exceed its current capacity.

See the Azure Disk Bursting documentation for more information about how bursting for Azure disks works.

If you regularly exceed your configured IOPS threshold, you can avoid depleting your burst credits by increasing your configured IOPS with any of the following actions:

  • Increase Cluster IOPS to increase the cluster's IOPS threshold.
  • Enable the Provision IOPS configuration option option. Clusters configured with this option use Provisioned IOPS SSD storage volumes, which do not use burst credits and offer greater IOPS thresholds. This option only applies to Atlas clusters on AWS.
  • Increase Storage Capacity. The IOPS threshold increases as you increase storage capacity. Clusters configured to use at least 1 TB of storage have baseline IOPS performance that is equal to or greater than the maximum burst performance. These volumes do not deplete burst credit balances.
  • Upgrade Cluster Tier. Larger cluster tiers include higher IOPS thresholds.

    Note

    Cluster tiers M140 and larger are deployed with at least 1 TB of storage capacity by default. Clusters with 1 TB or more of storage capacity do not deplete burst credit balances.

These are a few possible methods to observe high Disk I/O % Utilization:

  • The Util% graph in the cluster metrics displays a high value.
  • The disk IOPS use from the Disk IOPS graph in the cluster metrics exceeds the provisioned IOPS from the Atlas cluster configuration page.
  • The Normalized System CPU metric has a high IOWait curve. IOWait measures the percentage of time the CPU is idle and waiting for an I/O operation to complete. The Normalized System CPU chart is located under the Hardware Metrics section of the Metrics tab.
Give Feedback
MongoDB logo
© 2021 MongoDB, Inc.

About

  • Careers
  • Legal Notices
  • Privacy Notices
  • Security Information
  • Trust Center
© 2021 MongoDB, Inc.