Skip to main content
A newer release of this product is available.

Storage Node configuration settings

Contributors netapp-pcarriga
PDFs

Each Storage Node uses several configuration settings and counters. You might need to view the current settings or reset counters to clear alarms (legacy system).

Caution Except when specifically instructed in documentation, you should consult with technical support before modifying any Storage Node configuration settings. As required, you can reset event counters to clear legacy alarms.

Follow these steps to access a Storage Node's configuration settings and counters.

Steps

  1. Select SUPPORT > Tools > Grid topology.

  2. Select site > Storage Node.

  3. Expand the Storage Node and select the service or component.

  4. Select the Configuration tab.

The following tables summarize Storage Node configuration settings.

LDR

Attribute Name Code Description

HTTP State

HSTE

The current state of HTTP for S3, Swift, and other internal StorageGRID traffic:

  • Offline: No operations are allowed, and any client application that attempts to open an HTTP session to the LDR service receives an error message. Active sessions are gracefully closed.

  • Online: Operation continues normally

Auto-Start HTTP

HTAS

  • If selected, the state of the system on restart depends on the state of the LDR > Storage component. If the LDR > Storage component is Read-only on restart, the HTTP interface is also Read-only. If the LDR > Storage component is Online, then HTTP is also Online. Otherwise, the HTTP interface remains in the Offline state.

  • If not selected, the HTTP interface remains Offline until explicitly enabled.

LDR > Data Store

Attribute Name Code Description

Reset Lost Objects Count

RCOR

Reset the counter for the number of lost objects on this service.

LDR > Storage

Attribute Name Code Description

Storage State — Desired

SSDS

A user-configurable setting for the desired state of the storage component. The LDR service reads this value and attempts to match the status indicated by this attribute. The value is persistent across restarts.

For example, you can use this setting to force storage to become read-only even when there is ample available storage space. This can be useful for troubleshooting.

The attribute can take one of the following values:

  • Offline: When the desired state is Offline, the LDR service takes the LDR > Storage component offline.

  • Read-only: When the desired state is Read-only, the LDR service moves the storage state to Read-only and stops accepting new content. However, the LDR service continues to accept S3 or ILM driven purge and delete requests. Note that content might continue to be saved to the Storage Node for a short time until open sessions are closed.

  • Online: Leave the value at Online during normal system operations. The Storage State — Current of the storage component will be dynamically set by the service based on the condition of the LDR service, such as the amount of available object storage space. If space is low, the component becomes Read-only.

Health Check Timeout

SHCT

The time limit in seconds within which a health check test must complete in order for a storage volume to be considered healthy. Only change this value when directed to do so by Support.

LDR > Verification

Attribute Name Code Description

Reset Missing Objects Count

VCMI

Resets the count of Missing Objects Detected (OMIS). Use only after object existence check completes. Missing replicated object data is restored automatically by the StorageGRID system.

Verification Rate

VPRI

Set the rate at which background verification takes place. See information about configuring the background verification rate.

Reset Corrupt Objects Count

VCCR

Reset the counter for corrupt replicated object data found during background verification. This option can be used to clear the Corrupt Objects Detected (OCOR) alarm condition.

Delete Quarantined Objects

OQRT

Delete corrupt objects from the quarantine directory, reset the count of quarantined objects to zero, and clear the Quarantined Objects Detected (OQRT) alarm. This option is used after corrupt objects have been automatically restored by the StorageGRID system.

If a Lost Objects alarm is triggered, technical support might want to access the quarantined objects. In some cases, quarantined objects might be useful for data recovery or for debugging the underlying issues that caused the corrupt object copies.

LDR > Erasure Coding

Attribute Name Code Description

Reset Writes Failure Count

RSWF

Reset the counter for write failures of erasure-coded object data to the Storage Node.

Reset Reads Failure Count

RSRF

Reset the counter for read failures of erasure-coded object data from the Storage Node.

Reset Deletes Failure Count

RSDF

Reset the counter for delete failures of erasure-coded object data from the Storage Node.

Reset Corrupt Copies Detected Count

RSCC

Reset the counter for the number of corrupt copies of erasure-coded object data on the Storage Node.

Reset Corrupt Fragments Detected Count

RSCD

Reset the counter for corrupt fragments of erasure-coded object data on the Storage Node.

Reset Missing Fragments Detected Count

RSMD

Reset the counter for missing fragments of erasure-coded object data on the Storage Node. Use only after object existence check completes.

LDR > Replication

Attribute Name Code Description

Reset Inbound Replication Failure Count

RICR

Reset the counter for inbound replication failures. This can be used to clear the RIRF (Inbound Replication — Failed) alarm.

Reset Outbound Replication Failure Count

ROCR

Reset the counter for outbound replication failures. This can be used to clear the RORF (Outbound Replications — Failed) alarm.

Disable Inbound Replication

DSIR

Select to disable inbound replication as part of a maintenance or testing procedure. Leave unchecked during normal operation.

When inbound replication is disabled, objects can be retrieved from the Storage Node for copying to other locations in the StorageGRID system, but objects can't be copied to this Storage Node from other locations: the LDR service is read-only.

Disable Outbound Replication

DSOR

Select to disable outbound replication (including content requests for HTTP retrievals) as part of a maintenance or testing procedure. Leave unchecked during normal operation.

When outbound replication is disabled, objects can be copied to this Storage Node, but objects can't be retrieved from the Storage Node to be copied to other locations in the StorageGRID system. The LDR service is write-only.