Source Target Sizing

The Target Sizing section shows information about the source which can be used to choose the sizing of the target. Using information form the whole analysis and this section in particular can help in determining the size of a target to accommodate the current workloads.

The parameters in this section vary by source DB engine:

HBase

Parameter	Description
Table Count	Number of all user-defined tables in HBase.
Region Server Count	Number of all region servers that HMaster is currently aware of.
Region Count	Number of all regions across all user-defined tables.
Max. Used Heap Size	Size of the largest used heap on a single region server.
Avg. Used Heap Size	Average size of the used heaps across region servers.
Max. Requests Per Second	Largest number of requests per second on a single region server.
Avg. Requests Per Second	Average number of requests per second across region servers.
Max. Store File Index Size	Largest size of a store-file index on a single region server.
Avg. Store File Index Size	Average size of the store-file indexes across region servers.
Max. Uncompressed Store File Index Size	Largest size of an uncompressed store-file index on a single region server.
Avg. Uncompressed Store File Index Size	Average size of the uncompressed store-file indexes across region servers.
Max. Uncompressed Store File Size	Largest size of an uncompressed store-file on a single region server.
Avg. Uncompressed Store File Size	Average size of the uncompressed store-files across region servers.
Max. Store File Size	Largest size of a store-file on a single region server.
Avg. Store File Size	Average size of all store-files across region servers (excluding replication).
Max. Store File Count	Maximum number of store files on a single region server.
Avg. Store File Count	Average number of store files on a across region servers.
Total Uncompressed Store File Size	Sum of the uncompressed size of all store files across region servers (excluding replicas).
Total Store File Size	Sum of the size of all store files across region servers (excluding replicas).

MongoDB

Parameter	Description
Data node with max # CPU	The highest number of CPUs detected on any single data-bearing node.
Data node with max RAM (GB)	The highest RAM size detected on any single data-bearing node.
Number of shards	The number of shards in the cluster. On non-sharded clusters this value is zero.
Total collection count	The number of user collections across all data-bearing nodes. Only user-created collections are included in this count. Non sharded collections are counted in sharded collections. A sharded is counted as one, regardless of shard count. A replicated collection is counted as one, regardless of the number of replicas.
Total index count	The number of distinct indexes on all user collections across all data-bearing nodes.
Calibrate IO MBPS	Not applicable for MongoDB.
Multi-region setup	Indication of whether regional affinity was detected. This is based on tag-aware sharding setup. Applicable only to sharded clusters.
Max size on disk (GB)	The highest data size on disk detected on any single data-bearing node.
Number of users	Number of users across all databases. This includes the users defined in the `admin` database as well as those defined in user-created databases.
Number of custom roles	Number of roles user-defined (not built-in) roles, across all databases.
Max replica set member count	The maximum number of Replica Set members found in the cluster in a single Replica Set. In a sharded topology, this count relates to the shard with the highest number of members. In a single Replica Set topology, it is equivalent to the Replica Set’s member count. Not applicable for stand-alone deployment topology.

Oracle

Parameter	Description
Max average active sessions	Maximum across all active session averages collected.
Average 10 max active sessions	Average across top 10 session averages collected.
Average soft parse (%)	Ratio of soft parses vs. all statements parsed. A soft parse is one for which an existing reusable cached SQL statement exists.
Average CPU time ratio (%)	Average of database CPU time utilization.
Average max CPU time ratio (%)	Average of the database’s overall maximum CPU time utilization.
Average host CPU utilization (%)	Average CPU utilization by the host.
Average 50 host CPU utilization (%)	Average top 50 CPU utilization by the host.
Average Instance CPU Utilization (%)	Average percentage of time the instance processor is busy.
Avg Physical Read Ops/Sec	Physical Read IOPs per second.
Avg Physical Write Ops/Sec	Physical Write IOPs per second.
Peak redo/sec (MB)	Maximum changes per second based on the Redo log.
DB Block Changes Per Sec	DB block changes per second calculated over a maximum of the past 30 days.
Buffer cache hit (%)	Percentage of memory pages found in the buffer cache without having to read from disk. Based on cache hits divided over the past few thousand page-accesses statistics.
Max sessions	Maximum of the average number of sessions which were allowed detected in the database metrics. This includes both active and inactive sessions at those sampling periods.
Max active sessions	Maximum of the average number of active sessions detected in the database metrics.
Max allowed sessions	Configured limit on the number of concurrent sessions.
Calibrate IO IOPS	Maximum IOPS the source can handle.
Calibrate IO MBPS	Maximum MBPS the source can handle.
Physical to logical reads ratio (%)	Ratio of physical to logical reads. Physical reads are reads which could not be satisfied from PGA/cache and required a read from disk.
Redo Generated Per Sec (in MB)	Redo entries generated per second over a maximum of the past 30 days.
Physical Reads Total MBps	Source total physical reads per second in megabytes.
Physical Writes Total MBps	Source total physical writes per second in megabytes.
User Transaction Per Sec	User transaction per second over a maximum of the past 30 days.
User Commits Per Sec	User commits per second over a maximum of the past 30 days.
User Rollbacks Per Sec	User rollbacks per second over a maximum of the past 30 days.

Oracle RAC

Parameter	Description
Max average active sessions	Maximum across all active session averages collected.
Average 10 max active sessions	Average across top 10 session averages collected.
Average soft parse (%)	Ratio of soft parses vs. all statements parsed. A soft parse is one for which an existing reusable cached SQL statement exists.
Average CPU time ratio (%)	Average of database CPU time utilization.
Average max CPU time ratio (%)	Average of the database’s overall maximum CPU time utilization.
Average host CPU utilization (%)	Average CPU utilization by the host.
Average 50 host CPU utilization (%)	Average top 50 CPU utilization by the host.
Average Instance CPU Utilization (%)	Average percentage of time the instance processor is busy.
Avg Physical Read Ops/Sec	Physical Read IOPs per second.
Avg Physical Write Ops/Sec	Physical Write IOPs per second.
Peak redo/sec (MB)	Maximum changes per second based on the Redo log.
DB Block Changes Per Sec	DB block changes per second calculated over a maximum of the past 30 days.
Buffer cache hit (%)	Percentage of memory pages found in the buffer cache without having to read from disk. Based on cache hits divided over the past few thousand page-accesses statistics.
Max sessions	Maximum of the average number of sessions which were allowed detected in the database metrics. This includes both active and inactive sessions at those sampling periods.
Max active sessions	Maximum of the average number of active sessions detected in the database metrics.
Max allowed sessions	Configured limit on the number of concurrent sessions.
Calibrate IO IOPS	Maximum IOPS the source can handle.
Calibrate IO MBPS	Maximum MBPS the source can handle.
Physical to logical reads ratio (%)	Ratio of physical to logical reads. Physical reads are reads which could not be satisfied from PGA/cache and required a read from disk.
Redo Generated Per Sec (in MB)	Redo entries generated per second over a maximum of the past 30 days.
Physical Reads Total MBps	Source total physical reads per second in megabytes.
Physical Writes Total MBps	Source total physical writes per second in megabytes.
User Transaction Per Sec	User transaction per second over a maximum of the past 30 days.
User Commits Per Sec	User commits per second over a maximum of the past 30 days.
User Rollbacks Per Sec	User rollbacks per second over a maximum of the past 30 days.

PostgreSQL and MySQL

The fields shown for PostgreSQL and MySQL sources are:

Average 10 max active sessions
Average soft parse (%)
Average CPU time ratio (%)
Average max CPU time ratio (%)
Average host CPU utilization (%)
Average 50 host CPU utilization (%)
Buffer cache hit (%)
Calibrate IO IOPS
Calibrate IO MBPS
Max sessions
Max active sessions
Max allowed sessions
Max average active sessions
Peak redo/sec (MB)
Physical to logical reads ratio (%)

SQL Server

Parameter	Description
Max sessions	The maximum of the average number of sessions which were allowed detected in the database metrics. This includes both active and inactive sessions at those sampling periods.
Max active sessions	The average concurrent sessions among the top 10 connected session hosts.
Max allowed sessions	The configured limit on the number of concurrent sessions.
Buffer cache hit (%)	The percentage of memory pages found in the buffer cache without having to read from disk. Based on cache hits divided over the past few thousand page-accesses statistics.
Calibrate IO IOPS	Maximum IOPS the source can handle.
Calibrate IO MBPS	Maximum MBPS the source can handle.
Max average active sessions	The maximum of the average number of sessions which were allowed detected in the database metrics. This includes both active and inactive sessions at those sampling periods.
Average 10 max active sessions	Average across top 10 session averages collected.
Average soft parse (%)	The ratio of soft parses vs. all statements parsed. A soft parse is one for which an existing reusable cached SQL statement exists.
Average CPU time ratio (%)	Average of database CPU time utilization.
Average max CPU time ratio (%)	Average of the database’s overall maximum CPU time utilization.
Peak redo/sec (MB)	Maximum changes per second based on the Redo log.
Physical to logical reads ratio (%)	Ratio of physical to logical reads. Physical reads are reads which could not be satisfied from memory and required a hard read from disk.
Average host CPU utilization (%)	An estimated average of CPU utilization on the host, outside of the database.
Average 50 host CPU utilization (%)	An estimated average of CPU utilization on the host, outside of the database, over the top 50 utilization peaks.
Average Task Count	Average number of tasks currently running or scheduled to execute in the database.
Top Total Worker Time for entire instance	The maximum worker time detected for a single worker in a database across the instance.
Total I/O Utilization By source (MB)	Total I/O utilization across all databases. This measure includes reads and writes.
Top Waits 8	The highest wait time detected. Based on the maximum wait-duration detected.
Max Instance Memory (MB)	Maximum memory usage at the Instance level.
Resource Defined Limit (MB)	Maximum memory allowed to be utilized in the Instance.

Key Sizing Metrics

Applicable to Oracle and SQL Server only.

The Key Sizing Metrics section contains graphs that visualize the workload and target sizing information collected for the database source over time. The collected information is placed on a timeline, which allows zooming in/out to focus on specific time periods.

When applicable, graphs are complimented with the calculated value summary box providing the following information:

Key sizing metric and its calculated value, including an indicator whether it was calculated automatically or set manually.
Selected algorithm.
Any additional configured parameters, if applicable.

A cog icon in the upper-right corner of the calculated value summary box opens the Target Sizing section in Source Management.

When applicable, the following reference lines are placed on the graphs:

Maximum – the highest occurrence of all values.
99%, 95%, and 90% – the percentage of data points that are accounted for.
Average – the calculated average of all values.
Median – the value in the middle of the range found after sorting all values in ascending order.

Hovering over the timeline at any point highlights the value for the corresponding snapshot and the date for which the data was collected.

CPU tab.png — The **CPU** **Usage** graphs for an Oracle database

Oracle

The graphs represent the data for the focus period configured in Source Management. If changed, the graphs are updated respectively.

CPU

The CPU graphs display CPU utilization information over time.

AWR Repository

The following graphs are displayed:

Average Database CPU Usage – the average number of CPU threads used by the database during each snapshot time period.
Peak (non-sustained) Database CPU Usage – the peak number of CPU threads used by the database during each snapshot time period.

The Additional Metrics subsection contains the following graphs:

Comparison: Average CPU Usage, Database versus Entire Host – a comparison of the average number of CPU threads used for this database and the entire host.
Entire Host: Average CPU Usage – the average number of CPU threads used by the entire host during each snapshot time period.
Entire Host: Peak (non-sustained) CPU Usage – the peak number of CPU threads used by the entire host during each snapshot time period.

Statspack Repository

The following graph is displayed:

Peak (non-sustained) Database CPU Usage – the peak number of CPU threads used by the database during each snapshot time period.

Memory

The Memory graphs display memory allocation information over time.

AWR and Statspack Repositories

The following graph is displayed:

Memory Allocation – the combined allocated size in GB of the SGA and PGA.

IOPS

The IOPS graphs display data transfer rates information over time.

AWR Repository

The following graphs are displayed:

Average IOPS (IO Requests/second) – the average rate (Database IO Requests/second) during each snapshot time period.
Peak (non-sustained) IOPS (IO Requests/second) – the peak (non-sustained) rate (Database IO Requests/second) during each snapshot time period.

Statspack Repository

The following graph is displayed:

Peak (non-sustained) IOPS (IO Requests/second) – the peak (non-sustained) rate (Database IO Requests/second) during each snapshot time period.

IO Throughput

The IO Throughput graphs display data transfer rates information over time.

AWR Repository

The following graphs are displayed:

Average Throughput (MB/s) – the average IO Throughput rate (MB/s) during each snapshot time period.
Peak (non-sustained) Throughput (MB/s) – the peak (non-sustained) IO Throughput rate (MB/s) during each snapshot time period.

Statspack Repository

The following graph is displayed:

Peak (non-sustained) Throughput (MB/s) – the peak (non-sustained) IO throughput rate (MB/s) during each snapshot time period.

Disk Space

The Disk Space graphs display disk space utilization and trends within the database.

AWR Repository

The following graph is displayed:

Disk Space Metrics (GiB) over Time – the total allocated space of the non-TEMP tablespaces for the database.

Backup

The Backup graphs display backup-related information over time.

AWR Repository

The following graphs are displayed:

Redo Generation Rate (MB/s) over Time – the average redo generation rate (MB/s) during each snapshot time period.
Physical Write Rate (MB/s) over Time – the average write rate (MB/s) during each snapshot time period.

Each graph represents the data for the focus period set in Source Management. If changed, the graphs are updated respectively.

For Oracle RAC databases, graphs are displayed for the entire cluster. Instance-level graphs are found on the Single Instance page.

SQL Server

The graphs represent the data for the scan duration set in mMC. If changed, the graphs are updated respectively.

CPU

The CPU graphs display CPU utilization information over time.

The following graphs are displayed:

Average Database CPU Usage – the average CPU utilization calculated based on the total_worker_time metric collected using extended scan.
Peak Database CPU Usage – the peak CPU utilization calculated based on the total_worker_time metric collected using extended scan.

Memory

The Memory graphs display memory allocation information over time.

The following graph is displayed:

Memory Usage – the memory usage calculated based on the sys.dm_os_buffer_descriptors view.

IOPS

The IOPS graphs display data transfer rates information over time.

The following graphs are displayed:

Average IOPS (IO Requests/second) – the average IOPS calculated based on the sum of number of read and write operations collected using extended scan.
Peak (non-sustained) IOPS (IO Requests/second) – the peak IOPS calculated based on the sum of number of read and write operations collected using extended scan.

IO Throughput

The IO Throughput graphs display data transfer rates information over time.

The following graphs are displayed:

Average Throughput (MB/s) – the average IO Throughput calculated based on the sum of number of bytes read and written using extended scan.
Peak (non-sustained) Throughput (MB/s) – the peak IO Throughput calculated based on the sum of number of bytes read and written using extended scan.

Disk Space

The Disk Space graphs display disk space utilization and trends within the database.

The following graph is displayed:

Disk Space Metrics (GiB) over Time – the disk space in GiB calculated based on the sum of size over a specific period collected using extended scan.

Backup

The Backup graphs display backup-related information over time.

The following graph is displayed:

Backup Size – the backup size calculated based on the sum of backup size over a specific period, collected using extended scan.