Monitoring Your Volume Gateway

This section describes how to monitor a gateway in a cached volumes or stored volumes setup, including monitoring the volumes associated with the gateway and monitoring the upload buffer. You use the AWS Management Console to view metrics for your gateway. For example, you can view the number of bytes used in read and write operations, the time spent in read and write operations, and the time taken to retrieve data from the Amazon Web Services cloud. With metrics, you can track the health of your gateway and set up alarms to notify you when one or more metrics fall outside a defined threshold.

Storage Gateway provides CloudWatch metrics at no additional charge. Storage Gateway metrics are recorded for a period of two weeks. By using these metrics, you can access historical information and get a better perspective on how your gateway and volumes are performing. For detailed information about CloudWatch, see the Amazon CloudWatch User Guide.

Getting Volume Gateway Health Logs with Amazon CloudWatch Logs

You can use Amazon CloudWatch Logs to get information about the health of your Volume Gateway and related resources. You can use these logs to monitor your gateway for errors that it encounters. In addition, you can use Amazon CloudWatch subscription filters to automate processing of the log information in real time. For more information, see Real-time Processing of Log Data with Subscriptions in the Amazon CloudWatch User Guide.

For example, suppose that your gateway is deployed in a cluster activated with VMware High Availability (HA) and you need to know about any errors. You can configure a CloudWatch log group to monitor your gateway and get notified when your gateway encounters an error. You can either configure the group when you are activating the gateway or after your gateway is activated and up and running. For information about how to configure a CloudWatch log group when activating a gateway, see Configure your Volume Gateway. For general information about CloudWatch log groups, see Working with Log Groups and Log Streams in the Amazon CloudWatch User Guide.

For information about how to troubleshoot and fix these types of errors, see Troubleshooting volume issues.

The following procedure shows you how to configure a CloudWatch log group after your gateway is activated.

To configure a CloudWatch log group to work with your gateway

1. Sign in to the AWS Management Console and open the Storage Gateway console at https://console.aws.amazon.com/storagegateway/home.

2. In the left navigation pane, choose Gateways, and then choose the gateway that you want to configure the CloudWatch log group for.

3. For Actions, choose Edit gateway information, or on the Details tab, under Health logs and Not Enabled, choose Configure log group to open the Edit CustomerGatewayName dialog box.

4. For Gateway health log group, choose one of the following:

   • Disable logging if you don't want to monitor your gateway using CloudWatch log groups.

   • Create a new log group to create a new CloudWatch log group.

   • Use an existing log group to use a CloudWatch log group that already exists. Choose a log group from the Existing log group list.

5. Choose Save changes.

6. To see the health logs for your gateway, do the following:

   1. In the left navigation pane, choose Gateways, and then choose the gateway that you configured the CloudWatch log group for.

   2. Choose the Details tab, and under Health logs, choose CloudWatch Logs. The Log group details page opens in the Amazon CloudWatch console.

Using Amazon CloudWatch Metrics

You can get monitoring data for your gateway using either the AWS Management Console or the CloudWatch API. The console displays a series of graphs based on the raw data from the CloudWatch API. You can also use the CloudWatch API through one of the AWS Software Development Kits (SDKs) or the Amazon CloudWatch API tools. Depending on your needs, you might prefer to use either the graphs displayed in the console or retrieved from the API.

Regardless of which method you choose to use to work with metrics, you must specify the following information:

• The metric dimension to work with. A dimension is a name-value pair that helps you to uniquely identify a metric. The dimensions for Storage Gateway are GatewayId, GatewayName, and VolumeId. In the CloudWatch console, you can use the Gateway Metrics and Volume Metrics views to easily select gateway-specific and volume-specific dimensions. For more information about dimensions, see Dimensions in the Amazon CloudWatch User Guide.

• The metric name, such as ReadBytes.

The following table summarizes the types of Storage Gateway metric data that you can use.

CloudWatch Namespace	Dimension	Description
AWS/StorageGateway	GatewayId, GatewayName	These dimensions filter for metric data that describes aspects of the gateway. You can identify a gateway to work with by specifying both the GatewayId and the GatewayName dimensions. Throughput and latency data of a gateway are based on all the volumes in the gateway. Data is available automatically in 5-minute periods at no charge.
	VolumeId	This dimension filters for metric data that is specific to a volume. Identify a volume to work with by its VolumeId dimension. Data is available automatically in 5-minute periods at no charge.

Working with gateway and volume metrics is similar to working with other service metrics. You can find a discussion of some of the most common metrics tasks in the CloudWatch documentation listed following:

• Viewing Available Metrics

• Getting Statistics for a Metric

• Creating CloudWatch Alarms

Measuring Performance Between Your Application and Gateway

Data throughput, data latency, and operations per second are three measures that you can use to understand how your application storage that is using your gateway is performing. When you use the correct aggregation statistic, you can use Storage Gateway metrics to measure these values.

A statistic is an aggregation of a metric over a specified period of time. When you view the values of a metric in CloudWatch, use the Average statistic for data latency (milliseconds), use the Sum statistic for data throughput (bytes per second), and use the Samples statistic for input/output operations per second (IOPS). For more information, see Statistics in the Amazon CloudWatch User Guide.

The following table summarizes the metrics and corresponding statistic you can use to measure the throughput, latency, and IOPS between your applications and gateways.

Item of Interest	How to Measure
Throughput	Use the ReadBytes and WriteBytes metrics with the Sum CloudWatch statistic. For example, the Sum value of the ReadBytes metric over a sample period of 5 minutes divided by 300 seconds gives you the throughput as a rate in bytes per second.
Latency	Use the ReadTime and WriteTime metrics with the Average CloudWatch statistic. For example, the Average value of the ReadTime metric gives you the latency per operation over the sample period of time.
IOPS	Use the ReadBytes and WriteBytes metrics with the Samples CloudWatch statistic. For example, the Samples value of the ReadBytes metric over a sample period of 5 minutes divided by 300 seconds gives you IOPS.

For the average latency graphs and average size graphs, the average is calculated over the total number of operations (read or write, whichever is applicable to the graph) that completed during the period.

To measure the data throughput from an application to a volume

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. Choose Metrics, then choose the All metrics tab and then choose Storage Gateway.

3. Choose the Volume metrics dimension, and find the volume that you want to work with.

4. Choose the ReadBytes and WriteBytes metrics.

5. For Time Range, choose a value.

6. Choose the Sum statistic.

7. For Period, choose a value of 5 minutes or greater.

8. In the resulting time-ordered sets of data points (one for ReadBytes and one for WriteBytes), divide each data point by the period (in seconds) to get the throughput at the sample point. The total throughput is the sum of the throughputs.

The following image shows the ReadBytes and WriteBytes metrics for a volume with the Sum statistic. In the image, the cursor over a data point displays information about the data point including its value and the number of bytes. Divide the bytes value by the Period value (5 minutes) to get the data throughput at that sample point. For the point highlighted, the read throughput is 2,384,199,680 bytes divided by 300 seconds, which is 7.6 megabytes per second.

To measure the data input/output operations per second from an application to a volume

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. Choose Metrics, then choose the All metrics tab and then choose Storage Gateway.

3. Choose the Volume metrics dimension, and find the volume that you want to work with.

4. Choose the ReadBytes and WriteBytes metrics.

5. For Time Range, choose a value.

6. Choose the Samples statistic.

7. For Period, choose a value of 5 minutes or greater.

8. In the resulting time-ordered sets of data points (one for ReadBytes and one for WriteBytes), divide each data point by the period (in seconds) to get IOPS.

The following image shows the ReadBytes and WriteBytes metrics for a storage volume with the Samples statistic. In the image, the cursor over a data point displays information about the data point, including its value and the number of samples. Divide the samples value by the Period value (5 minutes) to get the operations per second at that sample point. For the point highlighted, the number of write operations is 24,373 bytes divided by 300 seconds, which is 81 write operations per second.

Measuring Performance Between Your Gateway and AWS

Data throughput, data latency, and operations per second are three measures that you can use to understand how your application storage using the Storage Gateway is performing. These three values can be measured using the Storage Gateway metrics provided for you when you use the correct aggregation statistic. The following table summarizes the metrics and corresponding statistic to use to measure the throughput, latency, and input/output operations per second (IOPS) between your gateway and AWS.

Item of Interest	How to Measure
Throughput	Use the ReadBytes and WriteBytes metrics with the Sum CloudWatch statistic. For example, the Sum value of the ReadBytes metric over a sample period of 5 minutes divided by 300 seconds gives you the throughput as a rate in bytes per second.
Latency	Use the ReadTime and WriteTime metrics with the Average CloudWatch statistic. For example, the Average value of the ReadTime metric gives you the latency per operation over the sample period of time.
IOPS	Use the ReadBytes and WriteBytes metrics with the Samples CloudWatch statistic. For example, the Samples value of the ReadBytes metric over a sample period of 5 minutes divided by 300 seconds gives you IOPS.
Throughput to AWS	Use the CloudBytesDownloaded and CloudBytesUploaded metrics with the Sum CloudWatch statistic. For example, the Sum value of the CloudBytesDownloaded metric over a sample period of 5 minutes divided by 300 seconds gives you the throughput from AWS to the gateway as bytes per second.
Latency of data to AWS	Use the CloudDownloadLatency metric with the Average statistic. For example, the Average statistic of the CloudDownloadLatency metric gives you the latency per operation.

To measure the upload data throughput from a gateway to AWS

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. Choose Metrics, then choose the All metrics tab and then choose Storage Gateway.

3. Choose the Gateway metrics dimension, and find the volume that you want to work with.

4. Choose the CloudBytesUploaded metric.

5. For Time Range, choose a value.

6. Choose the Sum statistic.

7. For Period, choose a value of 5 minutes or greater.

8. In the resulting time-ordered set of data points, divide each data point by the period (in seconds) to get the throughput at that sample period.

The following image shows the CloudBytesUploaded metric for a gateway volume with the Sum statistic. In the image, the cursor over a data point displays information about the data point, including its value and bytes uploaded. Divide this value by the Period value (5 minutes) to get the throughput at that sample point. For the point highlighted, the throughput from the gateway to AWS is 555,544,576 bytes divided by 300 seconds, which is 1.7 megabytes per second.

To measure the latency per operation of a gateway

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. Choose Metrics, then choose the All metrics tab and then choose Storage Gateway.

3. Choose the Gateway metrics dimension, and find the volume that you want to work with.

4. Choose the ReadTime and WriteTime metrics.

5. For Time Range, choose a value.

6. Choose the Average statistic.

7. For Period, choose a value of 5 minutes to match the default reporting time.

8. In the resulting time-ordered set of points (one for ReadTime and one for WriteTime), add data points at the same time sample to get to the total latency in milliseconds.

To measure the data latency from a gateway to AWS

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. Choose Metrics, then choose the All metrics tab and then choose Storage Gateway.

3. Choose the Gateway metrics dimension, and find the volume that you want to work with.

4. Choose the CloudDownloadLatency metric.

5. For Time Range, choose a value.

6. Choose the Average statistic.

7. For Period, choose a value of 5 minutes to match the default reporting time.

The resulting time-ordered set of data points contains the latency in milliseconds.

To set an upper threshold alarm for a gateway's throughput to AWS

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. Choose Alarms.

3. Choose Create Alarm to start the Create Alarm wizard.

4. Choose the Storage Gateway dimension, and find the gateway that you want to work with.

5. Choose the CloudBytesUploaded metric.

6. To define the alarm, define the alarm state when the CloudBytesUploaded metric is greater than or equal to a specified value for a specified time. For example, you can define an alarm state when the CloudBytesUploaded metric is greater than 10 MB for 60 minutes.

7. Configure the actions to take for the alarm state. For example, you can have an email notification sent to you.

8. Choose Create Alarm.

To set an upper threshold alarm for reading data from AWS

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. Choose Create Alarm to start the Create Alarm wizard.

3. Choose the StorageGateway: Gateway Metrics dimension, and find the gateway that you want to work with.

4. Choose the CloudDownloadLatency metric.

5. Define the alarm by defining the alarm state when the CloudDownloadLatency metric is greater than or equal to a specified value for a specified time. For example, you can define an alarm state when the CloudDownloadLatency is greater than 60,000 milliseconds for greater than 2 hours.

6. Configure the actions to take for the alarm state. For example, you can have an email notification sent to you.

7. Choose Create Alarm.

Understanding Volume Metrics

You can find information following about the Storage Gateway metrics that cover a volume of a gateway. Each volume of a gateway has a set of metrics associated with it.

Some volume-specific metrics have the same name as certain gateway-specific metrics. These metrics represent the same kinds of measurements but are scoped to the volume instead of the gateway. Before starting work, specify whether you want to work with a gateway metric or a volume metric. Specifically, when working with volume metrics, specify the volume ID for the storage volume that you want to view metrics for. For more information, see Using Amazon CloudWatch Metrics.

Note

Some metrics return data points only when new data has been generated during the most recent monitoring period.

The following table describes the Storage Gateway metrics that you can use to get information about your storage volumes.

Metric	Description	Cached Volumes	Stored Volumes
AvailabilityNotification	The number of availibility notifications sent by the volume. Units: count	Yes	Yes
CacheHitPercent	Percent of application read operations from the volume that are served from cache. The sample is taken at the end of the reporting period. When there are no application read operations from the volume, this metric reports 100 percent. Units: Percent	Yes	No
CachePercentDirty	The volume's contribution to the overall percentage of the gateway's cache that isn't persisted to AWS. The sample is taken at the end of the reporting period. Use the CachePercentDirty metric of the gateway to view the overall percentage of the gateway's cache that isn't persisted to AWS. For more information, see Understanding gateway metrics. Units: Percent	Yes	Yes
CachePercentUsed	The volume's contribution to the overall percent use of the gateway's cache storage. The sample is taken at the end of the reporting period. Use the CachePercentUsed metric of the gateway to view overall percent use of the gateway's cache storage. For more information, see Understanding gateway metrics. Units: Percent	Yes	No
CloudBytesDownloaded	The number of bytes downloaded from the cloud to the volume. Units: Bytes	Yes	Yes
CloudBytesUploaded	The number of bytes uploaded from the cloud to the volume. Units: Bytes	Yes	Yes
HealthNotification	The number of health notifications sent by the volume. Units: count	Yes	Yes
IoWaitPercent	The percentage of IoWaitPercent units that are currently used by the volume. Units: Percent	Yes	Yes
MemTotalBytes	The percentage of total memory that is currently used by the volume. Units: Percent	Yes	No
MemoryUsage	The percentage of memory that is currently used by the volume. Units: Percent	Yes	No
ReadBytes	The total number of bytes read from your on-premises applications in the reporting period. Use this metric with the Sum statistic to measure throughput and with the Samplesstatistic to measure IOPS. Units: Bytes	Yes	Yes
ReadTime	The total number of milliseconds spent on read operations from your on-premises applications in the reporting period. Use this metric with the Average statistic to measure latency. Units: Milliseconds	Yes	Yes
UserCpuPercent	The percentage of allocated CPU compute units that are currently used by the volume. Units: Percent	Yes	Yes
WriteBytes	The total number of bytes written to your on-premises applications in the reporting period. Use this metric with the Sum statistic to measure throughput and with the Samples statistic to measure IOPS. Units: Bytes	Yes	Yes
WriteTime	The total number of milliseconds spent on write operations from your on-premises applications in the reporting period. Use this metric with the Average statistic to measure latency. Units: Milliseconds	Yes	Yes
QueuedWrites	The number of bytes waiting to be written to AWS, sampled at the end of the reporting period. Units: Bytes	Yes	Yes