Known Issues in DataCore SANsymphony 10.0 PSP19

Asynchronous Replication

Creating a replication on a destination disk already served via loop back might fail and leave both sides on standby or only one side with successful replication. As a workaround, if only one side knows about the replication, perform a forced split and start over. If both are stating standby - activate the one that should be active.

Backups Using DataCore Windows Integration Kit (WIK) VSS Hardware Provider

Backup software running on Windows Hosts with WIK 4.1.2 (or earlier) may fail to create backups as expected when the VSS Hardware Provider component from WIK 4.1.2 is installed. This issue may prevent proper backup creation. As a workaround, use the native VSS provider from the Windows Host OS instead of the VSS Hardware Provider from WIK. Before using the VSS Hardware Provider for offloading snapshot creation as part of the host backup process, check the Release Notes of future WIK releases to verify that this issue is not identified as a Known Issue.

Capacity Optimization on Disk Pools

Capacity Optimization cannot be enabled on a disk pool configured with a sector size of 4KB. To support capacity optimization on disk pools, the ZFS driver creates Zvols with a sector size of 512B, which are added as physical disks to the disk pool. But the disk pool requires that all physical disks in the pool have the same sector size. As a workaround, configure the disk pool with a sector size of 512B to enable capacity optimization on the disk pool.

DataCore Management Console Issue

While upgrading from DataCore SANsymphony 10.0 PSP15 Update 3 or later, the DataCore Management Console (DMC) may crash if the DMC has the Snapshot Details and Settings tab open. This is non-critical and can be safely ignored. The DMC can be re-opened again normally as required.

Disk Pool Deletion

After deleting a disk pool, the DataCore Management Console may display an error: "Something went wrong.". As a workaround, restart the DataCore Management Console.

Disk Pool Mirroring

Disk pool mirroring may take longer to complete than expected while significant Host Write IO is being sent to the DataCore Server.

Disk Tier Value

An incorrect disk tier value “0” may display in the DataCore Management Console and the corresponding row in the allocation map may not display when the first physical disk in a pool mirror disk pair fails. This is a user interface issue and does not impact the auto-tiering operation. The disk tier value becomes unavailable for the time frame when the first physical disk in a pool mirror disk pair fails/becomes offline. This is a temporary issue and the correct tier value automatically displays once the failed/offline physical disk recovers or once the mirror set is split. A fix for this issue will be provided in a future release.

Encryption

When encrypted Continuous Data Protection (CDP) enabled or Sequential Storage-enabled virtual disks (vDisks) are imported with the encryption key to a different machine, disable the encryption on them. If re-imported to the parent machine with the encryption key, those vDisks will be in the failed state.

Initiator IP Ignored for iSCSI Connections

When configuring an iSCSI connection via either the DataCore Management Console or PowerShell Cmdlets, you can specify the initiator IP but it will be ignored.

Latency Thresholds

The custom maximum latency thresholds, configured before the software upgrade to the 10.0 PSP17 version remain active after the software upgrade and continue to produce alert notifications. The threshold definitions cannot be altered or removed after the upgrade. This may affect the following thresholds:

As a workaround, before upgrading the software to the 10.0 PSP17 version, reset all the maximum latency thresholds to default values. This removes the custom definitions from the configuration objects and allows the system to remove maximum latency alerting during the upgrade process as intended.

Users may not encounter this issue if the default threshold values have never been modified.

Log Recovery

Log recovery didn't start after upgrading backup server to PSP19 for Virtual Disk NWAY setup. The log recovery process can be restarted with the two following methods as workaround:

Workaround 1: Once the server is rebooted after the backup upgrade, manually stop and then start the updated server again.

Workaround 2: Follow these steps to restart the log recovery:

  1. On all nodes, navigate to C:\Program Files\DataCore\SANsymphony and run the “dcshadump /dump > outputfile.txt” command in administrative mode using Command Prompt. A new output file will be generated.

  2. In the output file, verify whether there is any line containing "IOsPaused = 1" (under the "TertiaryDevice" section) on all nodes. If you do not find any such lines, this workaround does not apply to your issue, and you should not continue.

  3. Run the dcscaptool.exe /dump DC.txt command (from the same C:\Program Files\DataCore\SANsymphony folder) and locate the WWN name of the updated node in the DC.txt file. For example, the WWN name in the DC.txt file is shown in the image below.

  4. Then, on all other nodes that have a mirror with the updated server, run the “dcscaptool /signalnodestart "updated node WWN" 1” command in administrative mode (from C:\Program Files\DataCore\SANsymphony). Example: “dcscaptool /signalnodestart 227D0030D9CAD4BE 1”.

  5. The log recovery will restart in approximately 2 minutes.

Mirrored Virtual Disks

On a three-way mirrored Virtual Disk (vDisk) configuration, the log recovery does not start following a replace or split-and-unserve on the mirror. As a workaround, rescan the ports.

Performance Recording

  • When recording a large number of counters on a local performance recording database, the database may become full and stale data may not be deleted as fast as new data is recorded. When this occurs, reading and recording performance data will stop until deletion operations are complete, and event log messages will be posted. Consider reducing the number of counters being recorded if this occurs.
  • Some counters may show differing values in recorded performance and live performance (Physical disk % Idle Time, Average Queue Length).

Purge on Pools with Mirrored Encrypted Virtual Disks

When a disk pool contains a mirrored encrypted virtual disk (vDisk), purging the failed disks in this pool may result in data corruption. Follow the steps below as a workaround:

  1. Disable "Auto Recovery" for the encrypted mirror vDisk.

  2. Purge the failed disk.

  3. Click "Replace" on the out-of-sync side.

  4. Enable "Auto Recovery" for the vDisk.

ReFS-formatted Virtual Disks

Formatting a virtual disk (vDisk) with Microsoft ReFS (Resilient File System) that is mapped over a loopback port will fail with an error stating the drive is not accessible.

Removing DataCore Servers from a Server Group

If after removing a DataCore Server from a server group, the removed server is still displayed in the DataCore Management Console, re-connect to the original server group by either using the ‘Server Group Connections’ Window or the ‘Connect to Server Group’ option in the Common Actions menu option.

Replication Transfer State

After exiting the test mode, if the device is not removed from the Windows Disk Management tool at the destination (standby) side, the Replication Transfer displays in the stuck state (the data was not transferring to the Replicated Virtual Disks (vDisks)) and few of the vDisks display an error such as destaging failure. As a workaround, manually refresh the Disk Management tool at the standby side.

Replication Issues in SANsymphony on Hyper-V Parent OS May Cause Standby Mode to Get Stuck

When SANsymphony is installed on the Hyper-V parent OS and asynchronous replication is configured for virtual disks, replication may get stuck in standby mode. To avoid this, install SANsymphony on a virtual machine (VM) instead of the parent OS.

Smart Deployment Wizard

  • The Smart Deployment Wizard (SDW) did not start the DataCore Executive Service (DCSX.exe) in the SDW-Clustered Virtual Machines Deployment Scenario after deploying the SANsymphony. As a workaround, check if the DCSX.exe is running in the target server. If not, start it manually and click "re-run" to continue the deployment in the SDW.
  • The Smart Deployment Wizard (SDW) failed to create the NFS share. Intermittently, SDW may fail to create the NFS share for the NAS Clustered configuration menu option and it does not report this failure. As a workaround, create the NFS share manually from Windows server>Failover cluster manager.

SMPA Disk Pools

Preserved space for an authorized SMPA disk pool may be reset to 0% when the authorized node changes. An example of this would be if the authorized node is stopped.

Snapshot Issues

  • The Events tab of the Snapshot in the DataCore Management Console may not display the operation completion message of a snapshot update procedure. However, the message does reflect in the DCS trace messaging and may be viewed by the support personnel if necessary. As a workaround, contact DataCore Technical Support for assistance with the trace messaging analysis.
  • When attempting to split a full snapshot, an error is displayed stating that the snapshots are not fully migrated as the migration map percentage is stuck. As a workaround, update the snapshot and retry splitting the snapshot.
  • A data miscompare may occur on a snapshot in a pool full scenario while executing "Promote" operation.

SNMP Output

After deleting the virtual disks (vDisks) and rebooting the servers, the name of some objects may display with extra spacing. This should be a non-functional problem as the object ID remains the same.

Virtual Disk Templates

Virtual disk templates of the 'Dual' type do not support Continuous Data Protection (CDP) or sequential storage.