Serialization Statistics
Introduction
The 'Serialization Statistics' Studio module provides information on how often data serialization is performed in Appway and shows the overall time that is spent on serializing and deserializing data.
To access the tool, go to System Maintenance > Statistics > Serialization Statistics.
Overview
In Appway, most data, such as Process data and Business Objects, is stored in distributed maps. These maps are partitioned among all nodes that participate in the cluster. Whenever a data item is written to a distributed map, Appway first serializes and compresses that item. Likewise, whenever a data item is read from a distributed map, that item is decompressed and deserialized.
The 'Serialization Statistics' table shows information in rows with a separate row for each map on every node in the cluster.
The data collected and displayed in the Serialization Statistics tool can help you to answer the following questions:
- How many CPU resources are used to serialize and deserialize data?
- Do I have data in my solution that is especially expensive to serialize and deserialize, such as too large value stores, or users with too many large attributes?
Refer to the following table for detailed descriptions of the Serialization Statistics columns:
| Column Name | Full Column Name | Description |
|---|---|---|
| Map | - | The map for which the statistics are displayed. One row for each map-node combination. |
| Persistence | - | Indicates if the data in this map is stored persistently. |
| Node | Node name | The name of the cluster node. The table contains a row for each map-node combination. |
| # OW | Number of 'object write' operations | The number of write operations performed for Java Serialization (for this map and node). |
| # OR | Number of 'object read' operations | The number of read operations performed for Java Serialization (for this map and node). |
| # DW | Number of 'data write' operations | The number of write operations initiated by Hazelcast (for this map and node). |
| # DR | Number of 'data read' operations | The number of read operations initiated by Hazelcast (for this map and node). |
| μ t(DW) | Average time per a 'data write' operation | The average time per operation spent on data writing (see # DW column) for this map and node. |
| μ° t(DW) | Rolling average time per a 'data write' operation | The average time that was spent per operation for the 1000 most recent data write operations (for this map and node). |
| μ t(DR) | Average time per a 'data read' operation | The average time per operation spent on data reading (see # DR column) for this map and node. |
| μ° t(DR) | Rolling average time per a 'data read' operation | The average time that was spent per operation for the 1000 most recent data read operations (for this map and node). |
| max s(U) | Maximum size uncompressed | The maximum uncompressed size of any data item for this map and node. |
| μ s(U) | Average size uncompressed | The average uncompressed size of all data items compressed for this map and node. |
| μ s(C) | Average size compressed | The average compressed size of all data items compressed for this map and node. |
| μ CR | Average compression ratio | The overall compression ratio (sum sizes compressed / sum sizes uncompressed) of all data items compressed (for this map and node). |
| μ t(C) | Average time compressing | The average time spent for compressing data (for this map and node). |
| μ t(D) | Average time decompressing | The average time spent for decompressing data (for this map and node). |
| # TLCC | Number of thread-local-cache clones | The number of clone operations for the thread-local cache (for this map and node). |
| μ t(TLCC) | Average time for thread-local cache clones | The average time per operation spent on clone operations (for this map and node). |
In addition to the rows showing the serialization statistics for all maps and nodes, the table contains aggregate rows with node name 'All'. These aggregate rows summarize the statistics for each map over all participating nodes.
As with other tables in the Appway Studio, you can add filter conditions at the top of some columns and you can select a column by which to sort the table contents.
Select Reset All Statistics (available in the context menu) to reset all serialization statistics.