Dell PowerStore 5200T EMC PowerStore Release Notes for PowerStore OS Version 3 - Page 13

Issue ID MDT-392418 MDT-387678 MDT-386942 MDT-386593 MDT-383302 MDT-366264 MDT-365446 MDT-362408 MDT-353803 MDT-350303 MDT-349512 Functional Area Description Install or Upgrade Install or Upgrade Install or Upgrade Install or Upgrade Install or Upgrade Install or Upgrade Install or Upgrade Install or Upgrade Install or Upgrade Install or Upgrade Install or Upgrade In rare conditions, a non-disruptive upgrade (NDU) may fail due to an unexpected High Availability Event caused by periodical hardware checks. Non-disruptive upgrade (NDU) fails due to device goes offline and comes back during NDU. PowerStore Node may keep rebooting and cannot enter service mode if a non-disruptive upgrade (NDU) fails during platform upgrade The state of the BBU in SYM has been observed to flip continuously between "ready" and "not ready". This results in excessive wear of NVMe drives as each time BBU goes to 'not ready' state the NVMe drives supported by that BBU will perform backup. Stale BBU information might cause NVRAM to go offline during node upgrade and nondisruptive upgrade (NDU) fails due to a disk event. After a failed upgrade results in data unavailability, in some instances, PowerStore Manager cannot be reached. A time-out occurred while waiting for commands to complete during the pre-upgrade health check causing the non-disruptive upgrade (NDU) to fail. After NDU failed, a rollback was not trigged, which blocked management operations. Before a non-disruptive upgrade (NDU), network issues between appliances can get some management DB commands not to be marked as "completed"/"failed" but stay as "pending"/"initial" state. By design, NDU fails to start if it finds that certain commands are in running state. Node B encounters a kernel panic due to blocked tasks running the internal partition manager. Due to known issues with AIX ACA handling, this causes conditions that affect other parts of the system to fail. This issue causes abort storms, session reestablishment, and other recovery processes that have additional negative side effects on the environment. An upgrade retry may fail if the previous non-disruptive upgrade (NDU) fails at the prepare stage and rollback does not happen. NVRAM disk bouncing during non-disruptive upgrade (NDU). MDT-347781 Install or Upgrade MDT-344080 Install or Upgrade MDT-335711 Install or Upgrade MDT-316869 Install or Upgrade MDT-315382 Install or Upgrade When a license is applied, the process may fail but the task may persist. This may impact an upgrade, which will fail indicating "upload license file to cluster" is running. Primary node cannot re-connect to a Peer Node during rollback after failed non-disruptive upgrade (NDU). Cluster initialization might fail due to a software module restart if the appliance is configured with eight cabled ports on the Fibre Channel IO module per node. Non-disruptive upgrade (NDU) may fail and lead to auto-recovered service interruption if the management port is down on one node during NDU. Pre-Upgrade Health Check (PUHC) may fail due to license file upload failure. MDT-299952 Install or Upgrade MDT-298855 Install or Upgrade MDT-254250 Install or Upgrade MDT-388750 Internal Migration A failed attempt to add an appliance to an existing cluster could result in PowerStore Manager being unavailable. Non-disruptive upgrade (NDU) fails due to an internal service failing to start, which leads to one node failing to join back the cluster. If a PowerStore appliance that is populated with more SSD drives than SCM drives is upgraded from PowerStore 1.0.3.x to PowerStore 2.0, the SCM drives might not be in use by the Data Path after the upgrade is complete. After ESXi host times out, the vVol migration session, which is between PowerStore appliances in the same cluster, may go into a System Paused state. Resolved Issues 13