Measures reported by PzStatMetaTest

As files are created and edited by users, files are written to the local Panzura CloudFS node they are working from, and CloudFS generates and adds metadata - information about the file.The file data and metadata are then deduplicated, compressed, encrypted, and split into 128KB blocks. Panzura physically decouples data from metadata, and integrates the global namespace into the metadata. This decoupling enables the file system to be highly flexible in referencing which physical blocks are used to construct a file. It also allows every node in the file system to hold a complete copy of the metadata for the whole file system, without having to hold the files themselves. Panzura CloudFS generates metadata pointers that reference the immutable objects that make up a file, and updates them in real time, every time a file changes. Read-only local and file system snapshots then capture the metadata pointers - and therefore the contents of the file system - at any given point in time. Metadata is stored centrally in the cloud for durability in addition to being fully cached locally for enhanced performance. All nodes in a single namespace or CloudFS synchronize metadata updates simultaneously every 60 seconds in a hub (cloud) and spoke (node) configuration. This is further complemented by a peer to peer (mesh) synchronization event that occurs in real-time when lock dynamically moves from one node to another through the distributed global file locking.

This test reports the statistics pertaining to the total and used number of metadata that are connected to the cloud controller being monitored. This way, administrators can keep a tab on the usage of metadata by the users of Panuzra Cloud file system.

Outputs of the test: One set of results for the target appliance being monitored.

The measures made by this test are as follows: