Issues Resolved in DataCore SANsymphony 10.0 PSP17 - Preview 1

3-Way Virtual Disk Host Access

While reverting a full snapshot to a backup (3rd) copy of a 3-way mirrored virtual disk (vDisk), the action was resulting in loss of host access to the vDisk.

Average Latency Operation

The DataCore Management Console always displayed the last calculated performance metrics for Read, Write, and Total Average Latencies even though the values were no longer increasing.

Auto-Tiering/Reserved Space Logging Issues

Excessive debug information about a Disk Pool’s ‘Reserved Space’ setting could fill the DataCore Service log file causing unnecessary log rollover and space usage on the C drive.

Configuration Issues

DataCore Servers were prevented from being added/ removed to/ from a Server Group or the DataCore Executive Service was unable to start up due to inconsistencies in shared configuration files. However, no obvious warning was displayed for the user.

DataCore Executive Service Issue

The DataCore Executive Service was crashing while collecting performance counters.

DataCore Management Console

The DataCore Management Console was reporting an “invalid cast exception” error when using the pass-through disks.

DataCore REST Installer

Installing or upgrading the DataCore REST component could cause an operating system restart midway through the installation or upgrade.

DataCore Server Issue

  • Adding a new virtual disk while monitoring DataCore's DcsSDCmapping counter using the Windows Performance Monitor was causing the DataCore Server to crash.
  • The DataCore Server either became 'unavailable' when viewed in the DataCore Management Console (DMC) or was unable to be logged in directly using the DMC if virtual disk mirrors were waiting to be synchronized at the same time as their mirror path mappings were being monitored for performance metrics. Also, the mirror recoveries did not start.

Encryption at Disk Pool

If both the capacity optimization and encryption features were enabled on a new virtual disk and the virtual disk's creation failed, the disk pool remained enabled for encryption even if it was not enabled before.

Failed SMTP Emails

Sending a test email using the SMTP settings was failing.

Incorrect DataCore Server License Expiry Date

DataCore Server was displaying an incorrect license expiry date when one of the servers in the server group was unavailable.

I/O Latency

During high load, SCSI UNMAP commands increased the latency for regular input/output (I/O).

Mirrored Virtual Disk

An internal timeout during the split-and-unserve operation was leaving a mirrored virtual disk in a permanently failed state.

Non-mirrored Virtual Disk Data Status

The VirtualDiskDataStatusMonitor always reported ‘Undefined’ for non-mirrored virtual disks regardless of their actual status.

Physical Disk

  • Purged offline, failed, or missing physical disks inside a disk pool were displayed as offline or failed disks in the DataCore Management Console even after successfully completing the purge disk wizard.
  • Creating and assigning physical disks to a disk pool using PowerShell failed with an error "The size of the pool will exceed the maximum possible size of 0 B".

REST Performance

The DataCore REST worker process was consuming all the available CPU resources on the monitoring server using Internet Information Services (IIS) for Windows, which impacted all other processes on the server.

SANsymphony Servers with Commvault Agent Installed

The DataCore Server crashed if a Commvault Agent was installed on or added to any SANsymphony servers running DataCore SANsymphony 10.0 PSP16 or PSP16 Update 1.

SCSI Resets

SCSI Resets in large configurations using mirrored virtual disks (vDisk) were resulting in unnecessary SCSI BUSY states until the RESETs were completed on both sides of the mirror. This led to increased IO latency or timeouts being reported on the hosts.

Sorting Virtual Disks

Sorting virtual disks on the “Allocated Space” and “Total Allocated Space” columns and sorting virtual disks in groups by these columns were not consistent.

Storage Allocation Unit Migrations

Disk pool physical disks that were already full were being considered a candidate for the storage allocation unit (SAU) migrations. This caused migration failures and led to a significant delay in the rate of all migrations within the disk pool.

System Managed Mirroring

If SANsymphony's System Managed Mirroring (SMM) feature was enabled then the virtual disk report created from either the DataCore Management Console or via the "Export-DcsReport" DataCore PowerShell commandlet was including internal-only device information that was not available to the user (also referred as Mirror Trunk Mappings).

VAAI Atomic Test & Set (ATS) Handling for ESXi Hosts

  • Atomic Test & Set (ATS) commands sent to a mirrored virtual disk that is served to a single ESXi Host returned undesirable SCSI BUSY status to the host.
  • Multiple ESXi hosts contending for the same ATS block saw frequent SCSI BUSY statuses returned if the hosts are configured with different preferred servers.

vCenter Integration

  • Bulk operations on ESXi hosts, such as creating and serving large numbers of virtual disks all at once, were failing.
  • When using SANsymphony's vCenter integration, the virtual disks were not discovered by the ESXi hosts immediately after being mapped.

Virtual Disk

  • A served virtual disk was unbrowsable on the host when it had a failed snapshot.
  • A served virtual disk configured as an asynchronous replication destination was not being discovered by a host after being made active or while in the Test Mode.

Virtual Disk ALUA State

The host's path to a mirrored virtual disk for a restarted DataCore Server was unavailable even after the successful completion of mirrored virtual disk synchronization.

Virtual Disk Deletion

The DataCore Server was crashing while deleting a mirrored virtual disk.