Symptoms

The system health check may display messages similar to: /cyc_node free space is less than 200 MB. (KB#000198173) Or Unable to parse passed url (0xE04040020008) The system may display a normal alert like: Root partition usage of node x has exceeded xx% on mount /cyc_node. Note: This alert or issue should not be ignored and must be resolved as soon as possible.If the system health checks detect this condition, PowerStoreOS upgrades may fail, and there is a risk of unexpected reboots.If the partition reaches 100%, SupportAssist may fail, and the node may enter Service Mode. In the following examples, both nodes of an appliance have insufficient space in /cyc_node. Running a System Check using PowerStore Manager with the Run Upgrade Extensions option Tip: The Profile field displays Upgrade Extensions when the Run Upgrade Extensions option is used to perform a System Check. Running a System Check using PowerStore Manager (UI) with the Run System Check option: Tip: The Profile field displays Service Engagement when the Run System Check option is used to perform a System Check. Running the System Health Service Script (svc_health_check) using an SSH connection: $ svc_health_check run -rhc cyc_node_space_check Name Location Status Description cyc_node space check node_a FAILED /cyc_node free space is less than 200MB. (KB#000198173) cyc_node space check node_b FAILED /cyc_node free space is less than 200MB. (KB#000198173) Example of the error: "Unable to parse passed url"

Cause

The above messages indicate that the node "/cyc_node" directory has insufficient space and requires attention. The "/cyc_node" can be full due to various reasons, for example: SupportAssist upload has too many accumulated failures due to firewall or SupportAssist backend issues. These may be introduced with firewall changes after SupportAssist is enabled, or might have another underlying issue that is related to firewall configuration.Large files are left over on /cyc_node as a result of other activities.PowerStore: Dell SupportAssist database may fill up and lead to a root partition full, causing either NDU failure or Node reboot This issue is partially resolved in PowerStoreOS 3.2.1.0.

Resolution

This issue was resolved in PowerStoreOS version 4.1 (or later). Workaround With earlier versions, if both nodes are in normal mode, you may attempt to temporarily disable SupportAssist. Usually, this clears out the SupportAssist files under "/cyc_node." Before attempting to disable SupportAssist, verify that the PowerStore is not exposed to PowerStore: Data Collection Fails, and the PowerStore Manager User Interface Reports Abnormal Information. If you suspect that the PowerStore may be experiencing this issue, resolve this first before proceeding to the next steps. Steps to disable SupportAssist from the PowerStore Manager user interface (UI) Settings > Support Connectivity > Connection Type > toggle to Disable and click Apply.Verify that SupportAssist did disable. Once the job appears to have finished, refresh the UI page and see if it shows disabled or enabled. If it shows enabled after a couple of attempts, then disabling SupportAssist from the UI did not work, go to step 5.If it does show disabled after a refresh of the UI page, continue to step 2. Wait at least 15 minutes. Reenable SupportAssist. Note: Even is enabling SupportAssist fails, it should be enough to reset this partition and clear the space. If SupportAssist fails to enable, normal troubleshooting should be carried out. Check /cyc_node partition on both nodes over SSH using the command: df -h If this does not resolve the issue, contact Dell Technical Support or your Authorized Service Representative, and reference this article ID. Note: This is only a temporary fix, and any underlying connectivity issues (firewall) must be fixed as well.

Support Cases