Skip to main content
Outside service Partner
Hitachi Vantara Knowledge

Monitoring replication


You can monitor the status of each currently configured replication link, as well as replication and recovery activity on the link, in the HCP System Management Console for either system involved in the link. By periodically reviewing link activity, you can decide whether to change the replication rate in the link configuration to accommodate other loads on system performance. You can also determine whether you need to add more storage capacity to either system.

Under normal circumstances, the replication service works without any intervention required. System Management Console alerts inform you of conditions, such as network connection problems, that may require action on your part to allow the service to continue processing.

For HCP tenants, you can control whether information about replication activity is displayed in the Tenant Management Console. For the default tenant, this information is always displayed.

This section of the Help describes the information available to you for the links you’ve created.

RoleWebHelp.png

Roles: To monitor replication, you need the monitor or administrator role.

NoteWebHelp.png

Note: You can also use the HCP management API to monitor replication links. For information on doing this, see Replication resources.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

Understanding the replication link list


The replication Links page in the HCP System Management Console lists the currently defined replication links in which the current system participates. To display this page, in the top-level menu of the HCP System Management Console, select Services Replication.

For each listed link, the replication Links page shows:

Name — The link name.

Type — The link type. The link type can be Active/Active, Outbound, or Inbound.

If the link is part of an erasure coding topology, the link type is followed by this icon: ErasureCodingEarth.png

Status — The link status, represented by an icon and text. The icon indicates the general health of the link. The text specifies the status of the link. The possible icons are:

o ReplicationStatusGood.png— The link is healthy. The statuses that can accompany this icon are:

Synchronizing data — On either system for an active/active link, the local system is sending data to the other system involved in the link.

Sending data — On the primary system for an active/passive link during replication, the primary system is sending data to the replica.

Receiving data — On the replica for an active/passive link during replication, the primary system is sending data to the replica.

o ReplicationStatusWarning.png — The link is healthy, but normal replication is not occurring on the link. The statuses that can accompany this icon are:

Recovering data — On the replica for an active/passive link during the first part of data recovery, the replica is sending data to the primary system.

Completing recovery — On the replica for an active/passive link during final data recovery, the replica is sending data to the primary system.

Receiving data — On the primary system for an active/passive link during data recovery, the replica is sending data to the primary system.

Scheduled off period: — No data is being sent on the link because the scheduled performance level for the current time period for the link is Off.

Suspended by user — On both systems for an active/active or active/passive link, activity on the link has been manually suspended. No replication or recovery is occurring.

o ReplicationStatusBad.png — The link is unhealthy. The statuses that can accompany this icon are:

Failed over — On both systems for an active/active or active/passive link, the link is currently failed over.

Remote storage full, link suspended — On the local system for an active/active link, on the primary system for an active/passive link during replication, or on the replica for an active/passive link during data recovery, the other system involved in the link does not have enough free space in primary running storage to accept any more data transmissions. In any case, HCP has automatically suspended activity on the link.

Local storage full, link suspended — On the local system for an active/active link, on the replica for an active/passive link during replication, or on the primary system for an active/passive link during data recovery, the local system does not have enough free space in primary running storage to accept any more data transmissions. In any case, HCP has automatically suspended activity on the link.

High error rate — On both systems for an active/active or active/passive link, errors are occurring at a high rate. If you see this status for:

An active/active link, check the health of both systems involved in the link

An active/passive link during replication, check the health of the replica.

An active/passive link during data recovery, check the health of the primary system

In any case, also check the health of the network connection between the two systems involved in the link.

If you cannot find the problem, contact your authorized HCP service provider for help.

Stalled link — On both systems for an active/active or active/passive link, activity on the link has stopped. Check the network connection between the two systems. If the connection appears to be working properly, contact your authorized HCP service provider for help.

Unrecognized link — On either system for an active/active or active/passive link, the other system involved in the link doesn’t recognize the link. Restore the link. Then, if applicable, start data recovery. For instructions on these activities, see Recovering from a failure.

Broken link — On both systems for an active/active or active/passive link, the replication service cannot contact any nodes on the other system involved in the link. If you see this status for:

An active/active link, check the health of both systems involved in the link

An active/passive link during replication, check the health of the replica

An active/passive link during data recovery, check the health of the primary system

In any case, also check the health of the network connection between the two systems involved in the link.

If you cannot find the problem, contact your authorized HCP service provider for help.

If more than one status applies to a link, the page shows the icon and text for only one of the statuses.

For additional information on the status of a link, hover over the status icon.

NoteWebHelp.png

Note: The System Management Console may not immediately reflect certain changes in link status.

Alerts — If replication or recovery activity is paused for one or more tenants on the link, either of these icons:

o WarnIconFilledIn.png — Replication or recovery has been manually paused for one or more tenants on the link.

o BadIconFilledin.png — Replication or recovery has been automatically paused for one or more tenants on the link. To view text describing the condition that’s causing this alert, hover over the alert icon.

For information on events that can cause replication or recovery of a tenant to be paused automatically, see Automatically paused tenant replication or recovery.

If replication or recovery is paused manually for one or more tenants and automatically for one or more other tenants, the page shows only the icon indicating an automatic pause.

The number of nodes on the other system that can receive replication transmissions from the current system. If the current system can determine the total number of storage nodes on the other system, the page shows that number as well.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

General link information


The list of links on the replication Links page shows the current status of each replication link. To further monitor the status of a link and to manage the link, click on the name of the link in the list.

The top part of the page that opens shows this information for the link:

The link name.

The name of the domain associated with the [hcp_system] network for the local system (the system where you’re viewing the link).

The name of the domain associated with the [hcp_system] network for the other system involved in the link (the remote system for the link).

A large arrow showing the direction of the link, overlaid with the link status. For an active/active link, this arrow has two arrowheads, one at each end.

For active/active links, the date and time before which configuration changes and changes to namespace content are guaranteed to be synchronized in both directions between the two systems involved in the link. For active/passive links, the date and time before which configuration changes and changes to namespace content are guaranteed to have been replicated or recovered on the link, as applicable.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

Link overview


To monitor the status of and activity on a replication link, you use the link status Overview panel in the HCP System Management Console. To display this panel:

1.In the list of links on the replication Links page, click on the link you want to monitor.

2.On the replication link details page, click on Status.

3.In the link Status panel, click on the Overview tab.

If the replication link is part of one or more active, retiring, or retired erasure coding topologies, the top part of the link status Overview panel lists the topologies that include that link.

The Sending and Receiving sections in the link status Overview panel show the current status of activity on the link:

For an active/active link, a section labeled Sending shows the status of send activity on the local system. A section labeled Receiving shows the status of send activity on the remote system.

For an outbound active/passive link during replication and an inbound active/passive link during data recovery, a section labeled Sending shows the status of send activity on the local system.

For an outbound active/passive link during data recovery and an inbound active/passive link during replication, a section labeled Receiving shows the status of send activity on the remote system.

The Sending and Receiving sections show:

Data pending — The approximate amount of data currently waiting to be sent to the other system, in bytes. This is the sum of the amounts of data waiting to be sent in each HCP namespace being replicated on the link. This value does not include data in the default namespace.

Information about pending data is not available in the Receiving section on a replica during replication or on a primary system during data recovery.

Objects pending — The approximate number of objects and parts of multipart objects currently waiting to be sent to the other system. This is the sum of the numbers of objects and object parts waiting to be sent in each HCP namespace being replicated on the link. This number does not include objects in the default namespace.

The number of objects pending does not include parts of in-progress multipart uploads. Those parts are not replicated until the applicable multipart upload has been completed.

Information about pending objects and object parts is not available in the section on a replica during replication or on a primary system during data recovery.

Up to date as of — The amount of time that is the difference between:

oThe date and time before which configuration changes and changes to namespace content are guaranteed to have been sent to the other system

oThe current date and time

If the replication verification service is running or has ever run, a Verifying section appears below the Receiving section. The Verifying section shows:

Last Pass Completed — The date and time at which the replication verification service finished its last run.

Status — The current status of the replication verification service. Possible values are:

oEnabled — The replication verification service is either running or set to run the next time the replication service runs.

oDisabled — The replication verification service is not running and is not set to run.

Issues Found — Either None or the current total number of nonreplicating objects and object parts in all the namespaces owned by the tenants on the link. If nonreplicating objects or object parts exist, View appears next to Issues Found. To see a list of the tenants with namespaces that contain nonreplicating objects or object parts, click on View. The list shows the total number of nonreplicating objects and ojbect parts for each tenant.

If you have administrative access to a listed tenant, you can click on the tenant name in the list to open the Tenant Management Console for that tenant. In the Tenant Management Console, you can see which namespaces have nonreplicating objects and object parts. For each of those namespaces, you can view a list of the nonreplicating objects and object parts in the namespace.

For more information on the replication verification service and nonreplicating objects and object parts, see Replication verification service processing.

The link status Overview panel also contains:

A graph showing the history of the rate of data transmissions on the link (see Data transmission rate)

A graph showing the history of the rate of operations on the link (see Operation rate)

For active/active links and outbound active/passive links, one or two pie charts showing the amount of used and free storage space on the remote system (see Remote storage space)

The statistics for replication links are updated every ten minutes. To see updated statistics, redisplay the Links page.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

Data transmission rate


The Transfer Rate graph in the link status Overview panel for a replication link shows the history of the rate of data transmissions on the link per second. For an active/active link, this is the rate of data transmission from the local system to the remote system. For an active/passive link, this is the rate of data transmission during replication or recovery, whichever is happening at the time. In any case, the data transmission rate is cumulative for all the HCP tenants and default-tenant directories on the link.

If the Transfer Rate graph is not currently visible, click on Transfer Rate to display it.

The x‑axis in the Transfer Rate graph marks the passage of time. The axis shows 30 days (or fewer if the replication link was created less than 30 days ago). The y-axis marks the data transmission rate in KB, MB, or GB. As the transmission rate varies, the measurement unit for the y-axis grows or shrinks as needed (for example, from KB to MB to GB).

The graph heading displays the current data transmission rate. Factors that affect this rate include the amount of other traffic on the network, the current load on both systems involved in the link, and whether data is being compressed or encrypted. Also, larger objects and object parts have higher rates of throughput.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

Operation rate


The Operations per Second graph in the link status Overview panel for a replication link shows the history of the rate of operations on the link per second. An operation is the replication of any of these:

An object, object part, directory, symbolic link, metadata change, or object deletion

An HCP tenant or HCP namespace or the modification or deletion of an HCP tenant or HCP namespace

For HCP tenants only, the creation, modification, or deletion of a user account

The creation, modification, or deletion of a retention class

The creation, modification, or deletion of a content class

A tenant log message

For an active/active link, the Operations per Second graph shows the operation rate for replication from the local system to the remote system. For an active/passive link, the graph shows the operation rate during replication or recovery, whichever is happening at the time. In any case, the operation rate is cumulative for all the tenants being replicated or recovered on the link.

If the Operations per Second graph is not currently visible, click on Operations per Second to display it.

The x‑axis in the Operations per Second graph marks the passage of time. The axis shows 30 days (or fewer if the replication link was created less than 30 days ago). The y-axis marks the operation rate in tens, hundreds, or thousands. As the operation rate varies, the measurement unit on the y-axis grows or shrinks as needed (for example, from tens to hundreds to thousands).

The graph heading displays the current rate of replication operations. Factors that affect the operation rate are the same as those that affect the data transmission rate.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

Remote storage space


The Remote Total Primary Storage pie chart in the link status Overview panel for an active/active link or an outbound active/passive link during replication shows the amounts of free and used space in primary storage for the other system involved in the link. Each amount is also shown as a number of bytes.

In a SAIN system with spindown storage, the storage being measured in this chart includes both primary running storage and primary spindown storage. In this case, a second pie chart to the right, labeled Remote Primary Running Storage, shows the amounts of free and used space in only primary running storage.

The two systems involved in a replication link do not necessarily have the same storage capacity or storage tiering strategies. If a system receiving replicated data starts to run low on free space in primary running storage, you need to evaluate whether to take action on that system to address the issue (for example, by increasing the primary running storage capacity or by tiering more data to other types of storage).

When primary running storage in the remote system for a replication link is almost full, the Overview page in the HCP System Management Console for the local system displays an alert. When the storage is full, the Console displays a different alert, and HCP automatically suspends replication on the link. For more information on these alerts, see System Management Console alerts.

For information on different types of storage, see Storage for HCP systems.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

Tenants view


To monitor the status of replication or recovery activity for the individual tenants on a given replication link, you use the link status Tenants panel in the HCP System Management Console. To display this panel:

1.In the list of links on the replication Links page, click on the name of the link you want.

2.On the replication link details page, click on Status.

3.In the link Status panel, click on the Tenants tab.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

Understanding the tenant list


The link status Tenants panel lists the tenants on a replication link, including the default tenant if the link includes any default-namespace directories. For each tenant, the list shows:

Name — The tenant name.

Sent as Of — For an active/active link, an outbound active/passive link during replication, or an inbound active/passive link during data recovery, the average of the sent-as-of times for the tenant's namespaces. The sent-as-of time for a namespace is the amount of time that is the difference between:

oThe date and time before which the namespace content is guaranteed to have been sent from the local system to the remote system

oThe current date and time

For the default tenant, this is the average sent-as-of time among the directories on the link.

TipWebHelp.png

Tips: 

If some tenants are significantly less up to date than others, set the replication priority for the link to Oldest Object First to reduce the gap.

If the sent-as-of time is consistently increasing for one or more tenants, even with the replication performance level set to High, you may need to add more processing power (for example, additional nodes) to your system or increase the bandwidth between the two systems involved in the link.

Received as Of — For an active/active link, an outbound active/passive link during data recovery, or an inbound active/passive link during replication, the average of the received-as-of times for the tenant's namespaces. The received-as-of time for a namespace is the amount of time that is the difference between:

oThe date and time before which the namespace content is guaranteed to have been sent from the remote system to the local system

oThe current date and time

For the default tenant, this is the average received-as-of time among the directories on the link.

Status — The current status of replication or recovery activity for the tenant, represented by one or more of these icons, as applicable:

o TenantReplicating.png — If the icon is animated, replication or recovery of the tenant is proceeding normally. If the icon is static, no replication or recovery is occurring on the replication link.

o TenantReplicationPaused.png — A user paused replication or recovery of the tenant.

o TenantReplicationAutopaused.png — The replication service automatically paused replication or recovery of the tenant. To view text describing the condition that’s causing this status, hover over the status icon.

For information on events that can cause replication or recovery of a tenant to be paused automatically, see Automatically paused tenant replication or recovery.

o TenantInChain.png — The tenant is on a link that’s chained into the link you’re viewing.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

Managing the tenant list


By default, the tenant list in the Tenants panel includes all tenants in the link. The tenants are listed 20 at a time in ascending order by tenant name.

You can page through, sort, and filter the list of tenants. The Tenants panel indicates which tenants are shown out of the total number of tenants in the current list.

Paging

You can change the number of tenants shown at a time in the Tenants panel. To do this, in the Items per page field, select the number of tenants you want. The options are 10, 20, and 50.

To page forward or backward through the tenant list, click on the next ( TenantForward.png ) or back ( TenantBack.png ) control, respectively.

To jump to a specific page in the tenant list:

1.In the Page field, type the page number you want.

2.Press Enter.

Sorting

You can sort the tenants in the tenant list by tenant name, outbound or inbound synchronization time, or status. To sort the list:

By name, click on the Name column heading. Each time you click on the column heading, the sort order switches between ascending and descending.

By outbound or inbound synchronization time, click on the Sent as Of or Received as Of column heading, as applicable. The first time you click on the column heading, the tenants are sorted in ascending order by the applicable synchronization time. Each time you click on the column heading after that, the sort order switches between ascending and descending.

By status, click on the Status column heading. The first time you click on the column heading, tenants for which replication has been automatically paused are listed first, followed by tenants for which replication has been manually paused, followed by the remaining tenants. Each time you click on the column heading after that, the order reverses from what it currently is.

Filtering

You can filter the tenant list by tenant name or tag. The filtered list includes only those tenants with a name or tag, as applicable, that begins with or is the same as a specified text string.

To filter the tenant list:

1.In the field above the Name column, select Name to filter by name or Tag to filter by tag.

2.In the next field, type the text string you want to use as a filter. This string can be up to 64 characters long, can contain any valid UTF-8 characters except commas (,), and is not case sensitive. White space is allowed.

3.Click on the find control ( FindControl.png ).

To redisplay the entire list of tenants after filtering it, click on the clear filter control ( ClearFilterControl.png ).

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

Tenant replication details


To see more details about the replication of a tenant, hover over the row for the tenant in the list of tenants in the link status Tenants panel. The information shown is:

Most up-to-date namespace:

oFor an active/active link, an outbound active/passive link during replication, or an inbound active/passive link during data recovery, the smallest sent-as-of time among the tenant's namespaces

oFor an outbound active/passive link during data recovery or an inbound active/passive link during replication, the smallest received-as-of time among the tenant's namespaces

For the default tenant, this is the smallest sent-as-of time or received-as-of time, as applicable, among the directories on the replication link.

Average up-to-date time:

oFor an active/active link, an outbound active/passive link during replication, or an inbound active/passive link during data recovery, the average of the sent-as-of times for the tenant's namespaces

oFor an outbound active/passive link during data recovery or an inbound active/passive link during replication, the average of the received-as-of times for the tenant's namespaces

For the default tenant, this is the average of the sent-as-of times or received-as-of times, as applicable, for the directories on the replication link.

Objects pending — For an active/active link, an outbound active/passive link during replication, or an inbound active/passive link during data recovery, the approximate number of objects and parts of multipart objects currently waiting to be sent to the remote system in the tenant's namespaces.

The number of objects pending does not include parts of in-progress multipart uploads. Those parts are not replicated until the applicable multipart upload has been completed.

Information about pending objects and object parts is not available for an outbound link during data recovery or for an inbound link during replication. This information is also not available for the default tenant.

Data pending — For an active/active link, an outbound active/passive link during replication, or an inbound active/passive link during data recovery, the approximate amount of data currently waiting to be sent to the remote system in the tenant's namespaces.

Information about pending data is not available for an outbound link during data recovery or for an inbound link during replication. This information is also not available for the default tenant.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

Controlling the Tenant Management Console replication display


For HCP tenants that are on at least one replication link, you control whether the Tenant Management Console displays information about the status of replication of the tenant and its namespaces. The Tenant Management Console for the default tenant always displays this information.

To control whether the Tenant Management Console displays replication status information for an HCP tenant and its namespaces:

1.In the top-level menu of the HCP System Management Console, select Services Replication.

2.On the left side of the Replication page, click on Settings.

3.On the replication Settings page, take either of these actions:

oTo have the Tenant Management Console display replication status information for all HCP tenants, select Allow tenants to monitor replication of their namespaces.

oTo have the Tenant Management Console hide replication status information for all HCP tenants, deselect Allow tenants to monitor replication of their namespaces.

4.Click on Update Settings.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.

System Management Console alerts


The Overview page in the HCP System Management Console has a section in which it displays alerts about abnormal conditions. Each alert consists of an icon and accompanying text that identifies the problem. Each alert also has text that’s displayed when you hover over the icon.

The table below describes the alerts that relate to replication. The alerts are listed alphabetically by their hover text.

Icon Mouse-over text Description
ErasureCodingErrorAlert.png Erasure coding error

One or more of these conditions exist:

The protection status of the active erasure coding topology is broken.

The read status of an erasure coding topology is broken.

One or more tenants added to an erasure coding topology have not been added to one or more replication links in the topology. HCP does not report this condition for a tenant until ten minutes after the tenant was added to the topology.

HCP cannot find one or more replication links that are included in an erasure coding topology.

ErasureCodingWarningAlert.png Erasure coding warning

One or more of these conditions exist:

The protection status of the active erasure coding topology is vulnerable.

The read status of an erasure coding topology is vulnerable.

Replication is paused for one or more tenants on one or more replication links in the active erasure coding topology.

The geo-distributed erasure coding service is not in the active service schedule. New erasure coding activity is limited.

The geo-distributed erasure coding service is disabled. New erasure coding activity is limited.

TrustedCertificateExpired.png

Expired replication certificates

One or more trusted replication server certificates have expired. Replication with the systems from which the applicable certificates were obtained has stopped.

When the other system in any of the applicable replication pairs installs a new SSL server certificate, download that certificate and upload it to this system as a trusted replication server certificate.

For ease of maintenance, delete expired certificates from the list of trusted replication server certificates.

ReplicationFrontEndNICError.png

Front-end connection error

All front-end connections to one or more nodes in the local system are unavailable. If the system is currently a source for replication, some objects and object parts will not be replicated. If the system is a replication target, replication performance is degraded.

IrreparableObjects.png Nonreplicating objects

One or more nonreplicating objects or object parts have been detected. These objects and object parts cannot be replicated until they are repaired. If the objects and object parts cannot be repaired, they are never replicated. If the system has nonreplicating objects or object parts that are irreparable, contact your authorized service provider for help.

PartialNetworkAssigned.png

Partial network assigned to replication

The network selected for replication is a partial network. The load from replication activity is not being fully distributed across the nodes in the system.

Check the replication Settings page to see which network is selected for replication. Then either assign IP addresses in those networks to all nodes, or select a different network for replication.

ReplicaCapacityWarning.png

Remote storage almost full

Primary running storage space on the remote system for a replication link is at least 90% used. Consider adding more primary running storage capacity to the remote system or changing storage tiering strategies on that system to create more space in the existing primary running storage.

For an active/active link, this alert appears only in the System Management Console for the local system. For an active/passive link, this alert appears only in the System Management Console for the primary system.

ReplicaCapacityBad.png

Remote storage full (link suspended)

Primary running storage space on the remote system for a replication link is 94% used. The remote system cannot act on any more replication data transmissions from the local system. HCP has automatically suspended activity on the link.

Either add more primary running storage capacity to the remote system, or change storage tiering strategies on that system to create more space in the existing primary running storage. Then resume activity on the link.

For an active/active link, this alert appears only in the System Management Console for the local system. For an active/passive link, this alert appears only in the System Management Console for the primary system.

TrustedCertificateWarning.png

Replication certificates expire soon

One or more trusted replication server certificates will expire within 90 days. If a certificate expires, replication with the system from which the certificate was obtained will fail.

To ensure that replication is not disrupted, when the other system in the replication pair installs a new SSL server certificate, download that certificate and upload it to this system as a trusted replication server certificate.

TrustedCertificateCritical.png

Replication certificates expire soon

One or more trusted replication server certificates will expire within 30 days. If a certificate expires, replication with the system from which the certificate was obtained will fail.

To ensure that replication is not disrupted, when the other system in the replication pair installs a new SSL server certificate, download that certificate and upload it to this system as a trusted replication server certificate.

replication-link-broken.png

Replication link failure

An active/active or outbound active/passive replication link is not working as expected. Check the network connection between the two systems involved in the link. If the connection appears to be working properly, contact your authorized HCP service provider for help.

ReplicationAutopausedTenantsAlert.png

Replication link has autopaused tenants

HCP has automatically paused replication of one or more HCP tenants. For each tenant, correct the situation that caused replication to be paused. Then resume replication of the tenant.

replication-link-pending.png

Replication link pending

Either the system has sent a request for a replication link to another system and is waiting for a response, or the system has received a request for a replication link and has not yet responded.

replication-link-stalled.png

Replication link stalled

Replication has unexpectedly stopped on a replication link. Check the network connection between the two systems involved in the link. If the connection appears to be working properly, contact your authorized HCP service provider for help.

ReplicationLinksShutDown.png

Replication links shut down

All replication links in which this system is involved are shut down. No replication or recovery activity is occurring on these links, and the links cannot be used for read-from-remote or object-repair purposes.

TimeOutOfSync.png

Time out of sync between replicating systems

The system time on this system is more than one minute out of sync with the system time on one or more other systems with which this system participates in a replication link.

The recommended practice is to have all HCP systems in a replication topology use the same external time server.

For more information on alerts, see Alerts.

© 2015, 2019 Hitachi Vantara Corporation. All rights reserved.