Problem description
An administrator may observe an alert with the message: "CloudCache unable to read/write on Volume – Drive:\Phoenix\PhoenixCacheStore".
An examination of the diagnostic logs provides further insight. The Phoenix log file often contains a MemoryError traceback, while the Windows Event Viewer typically displays a resource exhaustion event, specifically Event ID 2004, pinpointing a low virtual memory condition on the server.
Cause
This issue is most frequently caused by the server running out of available virtual memory. However, other potential root causes include:
Connectivity Issues: The drive containing the cache is detached, or the entire server is offline.
Hardware Failure: A problem exists with the server's underlying RAID controller.
Traceback
Log location: C:\ProgramData\PhoenixCloudCache\Phoenixcloudcache.log
[2020-01-31 06:25:08,348] [INFO] [wpid 943-3568-1580468901] CacheFlush activity disconnected. wid 1
[2020-01-31 06:25:08,348] [ERROR] Error <type 'exceptions.MemoryError'>:. Traceback -Traceback (most recent call last):
Event log
event id 2004.
The Resource Exhaustion event.
Windows successfully diagnosed a low virtual memory condition.
The following programs consumed the most virtual memory: <Application1>.exe (10204) consumed 480612352 bytes, <Application2>.exe (2372) consumed 164450304 bytes, and PhoenixCacheServer.exe (5796) consumed 96604160 bytes
Resolution
The resolution involves mitigating the memory pressure on the server. The following steps should be taken:
Manage High-Consumption Applications: The administrator should identify the top memory-consuming applications listed in the Event ID 2004 log. Non-essential or problematic applications should be managed or closed to free up system resources.
Increase Virtual Memory: As an immediate measure, the server's virtual memory (page file) size can be increased. This provides a temporary buffer and can help stabilize the system.
Increase Physical Memory (RAM): For a long-term and effective solution, it is recommended to install additional physical RAM in the server. This directly addresses the resource bottleneck and prevents the error from recurring.
Note : If the read/write errors are frequent, as a part of immediate action unmap the csets from the Cloud-cache for time being until you perform the troubleshooting mentioned in the resolution part.