{{filterText}}
\n{{filterText}}
\nClick here to specify custom Date Range for chart. Custom Date Range in effect.
\nClick here to specify custom Date Range for chart. Custom Date Range in effect.
\n[AWS White Paper] Migrating Oracle Exadata Workloads to AWS
\n\n[AWS Guidance] Blueprint for successful migrations from Oracle Exadata to AWS
\n\n \n\n[Customer Blog - Paysafe - part1] Migrating Exadata workloads to AWS
\n[part2] Migrating Exadata workloads to AWS, the RDS service
\n[part3] Migrating Exadata workloads to AWS, hard-core EC2 services
\n[part4] Migrating Exadata workloads to AWS, using cloud-native data stores
\n[part5] Migrating Exadata workloads to AWS, EBS volumes for Oracle DBAs
\n[part6] Predicting storage requirements
\n \n[Customer Video - CalHEERS] Large-scale, Oracle Engineered Systems migration to the cloud
\n \n \n[Partner Blog - TCS] Comparing Oracle Exadata Database Performance with Amazon RDS for Oracle
\n[Partner Blog - Navisite] Proving the Performance of Oracle Exadata-Based Workloads on AWS
\n[Partner Blog - Tekstream] Migrating On-Premises Oracle Exadata Databases to Amazon RDS
\n[Partner White Paper - Apps Associates] Modernize your Exadata Workloads with AWS
\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n\n\n\n\nThese charts show the percentile distribution of various metrics (such as Database Size) for databases analyzed by DBCSI. These charts show you how your database's metrics compare relative to other databases. The horizontal axis is the metric displayed using a logarithmic scale. The vertical axis represents the percentile for databases ordered by the metric. 0 on the vertical axis is the smallest database for that metric. 50 on the vertical axis represents the database with the median value. 100 on the vertical axis represents the largest database for that metric. Each database is plotted as a dot showing its metric value and its percentile.
\n\n {{item.version}}, {{item.total_size}}\n \n
\n {{item.descr}}\n
\n Upload Date: {{item.collection_date}}\n
\n Upload User: {{item.upload_user}}\n
\n Status: {{item.status}}\n
{{filterText}}
\nClick here to specify custom Date Range for chart. Custom Date Range in effect.
\nThis feature lets you share your DBCSI databases via generated obscure URLs, similar to how you share in the AWS Pricing Calculator. The other users who wish to use your generated sharing URL will need to be DBCSI users as well. The generated URL does not copy over your algorithm choices or any manually entered target values. To continue, please read and acknowledge the following:
\nSharing Acknowledgement: Any DBCSI user who has been given the generated sharing URL will be able to see the DBCSI information for this database.
\n \n\nAre you sure you want to delete this database from DBCSI?
\nThese charts show the percentile distribution of various metrics (such as Database Size) for databases analyzed by DBCSI. These charts show you how your database's metrics compare relative to other databases. The horizontal axis is the metric displayed using a logarithmic scale. The vertical axis represents the percentile for databases ordered by the metric. 0 on the vertical axis is the smallest database for that metric. 50 on the vertical axis represents the database with the median value. 100 on the vertical axis represents the largest database for that metric. Each database is plotted as a dot showing its metric value and its percentile.
\nThis shows the average # of CPU threads used by the database during each snapshot time period using the Oracle metric: (CPU Usage Per Sec). It can be greater than 1 when multiple CPU threads at active at the same time. For a RAC cluster, this chart aggregates the usage across all instances. Use this chart to help identify a target baseline of CPU threads needed for this database.
\n\n This shows the peak # of CPU threads used by the database during each snapshot time period using the Oracle metric: (CPU Usage Per Sec). It can be greater than 1 when multiple CPU threads at active at the same time. For a RAC cluster, this chart aggregates the usage across all instances. Use this chart, along with the above Average Database CPU Usage chart, to help identify a target baseline of CPU threads needed for this database. Be sure to run the latest DBCSI data collection SQL script if you do not see any data.
\n\n This compares the average # CPU threads used for this database versus the entire host. It is comparing these Oracle metrics: (CPU Usage Per Sec, Host CPU Per Sec). This chart is mostly for informational purposes-- use this chart to get an understanding if other databases or other programs are running on the same host (or hosts in the case of an Oracle RAC cluster). This is a non-stacked chart.
\n\n This shows the average # of CPU threads used by the entire host during each snapshot time period. This is the Oracle metric: (Host CPU Usage Per Sec). It can be greater than 1 when multiple CPU threads at active at the same time. For a RAC cluster, this chart aggregates the usage across all instances. This chart is mostly for informational purposes-- typically you would use the Database CPU Usage charts to determine your target baseline of CPU threads needed for the database.
\n\n This shows the peak # of CPU threads used by the entire host during each snapshot time period. This is the Oracle metric: (Host CPU Usage Per Sec). It can be greater than 1 when multiple CPU threads at active at the same time. For a RAC cluster, this chart aggregates the usage across all instances. This chart is mostly for informational purposes-- typically you would use the Database CPU Usage charts to determine your target baseline of CPU threads needed for the database. Be sure to run the latest DBCSI data collection SQL script if you do not see any data.\n
\nBecause collecting metrics from PDBs can be difficult, we have collected CPU metrics from a 2nd data source so that you can compare it. This chart shows the average # of CPU threads used by the database during each snapshot time period using the Oracle metric: (CPU_CONSUMED_TIME from DBA_HIST_RSRC_PDB_METRIC). Values at the PDB-level are sanity-checked against the CDB (if found in DBCSI) or against typical thresholds, and outliers are not shown. It can be greater than 1 when multiple CPU threads at active at the same time. For a RAC cluster, this chart aggregates the usage across all instances. Use this chart to help identify a target baseline of CPU threads needed for this database.
\n\n This shows the combined allocated size in GB of the SGA and PGA. For Oracle RAC clusters, the sizes are the aggregate across all instances. Use this chart to help select a target memory configuration for this database (be sure to allow for some additional memory for the operating system). This is a stacked chart.
\nThe below chart shows the average IOPS during each snapshot time period using the Oracle metric: (IOPS from DBA_HIST_RSRC_PDB_METRIC). This is a PDB-specific metric and may not include some IOPS that happen at the CDB-level. For a RAC Cluster, the chart aggregates the metrics across all the instances. Use this chart to help select a target baseline for IOPS for this database.
\n\n This shows the average rate (Database IO Requests/second) during each snapshot time period using the Oracle metrics: (Physical Read Total IO Requests Per Sec, Physical Write Total IO Requests Per Sec). For a RAC Cluster, the chart aggregates the metrics across all the instances. Use this chart to help select a target baseline for IOPS for this database. This is a stacked chart.\n
\n\n This shows the peak (non-sustained) rate (Database IO Requests/second) during each snapshot time period using the Oracle metrics: (Physical Read Total IO Requests Per Sec, Physical Write Total IO Requests Per Sec). For a RAC Cluster, the chart aggregates the metrics across all the instances. Use this chart, along with the above Average IOPS chart, to help select a target baseline for IOPS for this database. This is a stacked chart.\n
\n\n This shows the average IO throughput rate (MB/s) during each snapshot time period using the Oracle metrics: (Physical Read Total Bytes Per Sec, Physical Write Total Bytes Per Sec). For a RAC Cluster, the chart aggregates the metrics across all the instances. For an Exadata system, this does include I/O from the storage cells even I/O avoided via storage index use. Use this chart to help select a target baseline for Throughput for this database. This is a stacked chart.\n
\nBecause collecting metrics from PDBs can be difficult, we have collected IO Throughput metrics from a 2nd data source so that you can compare it. This chart shows the average IO throughput rate (MB/s) during each snapshot time period using the Oracle metric: (IOMBPS from DBA_HIST_RSRC_PDB_METRIC). Values at the PDB-level are sanity-checked against the CDB (if found in DBCSI) or against typical thresholds, and outliers are not shown. For a RAC Cluster, the chart aggregates the metrics across all the instances. For an Exadata system, this does include I/O from the storage cells even I/O avoided via storage index use. Use this chart to help select a target baseline for Throughput for this database.
\n\n This shows the peak (non-sustained) IO throughput rate (MB/s) during each snapshot time period using the Oracle metrics: (Physical Read Total Bytes Per Sec, Physical Write Total Bytes Per Sec). For a RAC Cluster, the chart aggregates the metrics across all the instances. For an Exadata system, this does include I/O from the storage cells even I/O avoided via storage index use. Use this chart, along with the above Average Throughput chart, to help select a target baseline for Throughput for this database. This is a stacked chart.
\nYou can use the following controls if you want to narrow down the date range shown in the various charts on this page.\n
\n Note: The custom date range you specify below only affects the data shown in the charts. The auto sizing suggestion algorithms will continue to use the full date range when calculating target values. But you can use the custom date range with charts as a tool to help when you manually pick your own target values.
You can use the following controls if you want to narrow down the date range shown in the various charts on this page.\n
\nAdditional Preferences
\n\n\n{{ ((row.item.IS_PDB=='YES')) ? \"This shows the average # of CPU threads used by the PDB during each snapshot time period using the Oracle metric: (CPU Usage Per Sec from DBA_HIST_CON_SYSMETRIC_SUMM) along with the alternate metric: (CPU_CONSUMED_TIME from DBA_HIST_RSRC_PDB_METRIC). Values at the PDB-level are sanity-checked against the CDB (if found in DBCSI) or against typical thresholds, and outliers are not shown. The value can be greater than 1 when multiple CPU threads at active at the same time. For a RAC cluster, this chart aggregates the usage across all instances. Use this chart to help identify a target baseline of CPU threads needed for this database.\" : \"This shows the average and peak # of CPU threads used by the database during each snapshot time period using the Oracle metric: (CPU Usage Per Sec). The value can be greater than 1 when multiple CPU threads at active at the same time. For a RAC cluster, this chart aggregates the usage across all instances. Use this chart to help identify a target baseline of CPU threads needed for this database.\" }}
\n \n 0)\">The following chart has been adjusted for AIX CPU Usage timekeeping by using an AIX Power SMT CPU Time Adjustment multiplier of {{row.item.AIX_CPU_ADJUSTMENT_USED}}. \n The Average and Peak CPU Usage lines have been adjusted from the source system's threads ({{ row.item.PLATFORM_NAME }}) to x86 VCPUs using the CPU Thread Ratio of {{row.item.cpu_thread_ratio ?? form.cpu_thread_ratio}}.\n{{ row.item.ExadataVCPUAdjustmentText }}
\nAnd the below chart is the breakout of the Average DB CPU Usage by RAC Instance. This chart can show you how the workload is spread across the different RAC nodes.
\nAnd the below chart is the breakout of the Average DB CPU Usage by child database. This chart can show you how the workload is spread across the different child databases being consolidated together. This is a stacked chart.
\n{{ ((row.item.IS_PDB=='YES')) ? \"This shows the average IO throughput rate (MB/s) during each snapshot time period using the Oracle metrics: (Physical Read Total Bytes Per Sec, Physical Write Total Bytes Per Sec from DBA_HIST_CON_SYSMETRIC_SUMM) along with the alternate metric: (IOMBPS from DBA_HIST_RSRC_PDB_METRIC). Values at the PDB-level are sanity-checked against the CDB (if found in DBCSI) or against typical thresholds, and outliers are not shown. For a RAC Cluster, the chart aggregates the metrics across all the instances. For an Exadata system, this does include I/O from the storage cells even I/O avoided via storage index use. Use this chart to help select a target baseline for Throughput for this database.\" : \"This shows the average and peak IO throughput rate (MB/s) during each snapshot time period using the Oracle metrics: (Physical Read Total Bytes Per Sec, Physical Write Total Bytes Per Sec). For a RAC Cluster, the chart aggregates the metrics across all the instances. For an Exadata system, this does include I/O from the storage cells even I/O avoided via storage index use. Use this chart to help select a target baseline for Throughput for this database.\" }}
\nThe below chart is the breakout of the Average Throughput by child database. This chart can show you how the workload is spread across the different child databases being consolidated together. This is a stacked chart.
\nThe below chart shows the estimated I/O Throughput after adjusting for post-Exadata tuning activities {{row.item.ExadataAdjustmentText}}.
\nThe adjustment logic is to start with the target value for I/O throughput as seen on the source Exadata ({{row.item.OrigThroughput}} MB/s). The {{(row.item.smartscan_algo=='ss2') ? 'MODERATE' : ((row.item.smartscan_algo=='ss3') ? 'HEAVY' : ((row.item.smartscan_algo=='ss4') ? 'VERY HEAVY' : 'LIGHT'))}} tuning assumption assumes that I/O Throughput can be reduced by a factor of {{(row.item.smartscan_algo=='ss2') ? '3' : ((row.item.smartscan_algo=='ss3') ? '5' : ((row.item.smartscan_algo=='ss4') ? '10' : '2') )}}. More discussion available internally here.
\n \n\nThis shows the average IOPS during each snapshot time period using the Oracle metric: (IOPS from DBA_HIST_RSRC_PDB_METRIC). This is a PDB-specific metric and may not include some IOPS that happen at the CDB-level. For a RAC Cluster, the chart aggregates the metrics across all the instances. Use this chart to help select a target baseline for IOPS for this database.
\nThis shows the average and peak rate (Database IO Requests/second) during each snapshot time period using the Oracle metrics: (Physical Read Total IO Requests Per Sec, Physical Write Total IO Requests Per Sec). For a RAC Cluster, the chart aggregates the metrics across all the instances. Use this chart to help select a target baseline for IOPS for this database.
\nThe below chart is the breakout of the Average IOPS by child database. This chart can show you how the workload is spread across the different child databases being consolidated together. This is a stacked chart.
\nThe below chart shows the estimated IOPS after adjusting for post-Exadata tuning activities {{row.item.ExadataAdjustmentText}}.
\nThe adjustment logic is to start with the target value for IOPS as seen on the source Exadata ({{row.item.OrigIOPS}}). The {{(row.item.smartscan_algo=='ss2') ? 'MODERATE' : ((row.item.smartscan_algo=='ss3') ? 'HEAVY' : ((row.item.smartscan_algo=='ss4') ? 'VERY HEAVY' : 'LIGHT'))}} tuning assumption assumes that IOPS can be reduced by a factor of {{(row.item.smartscan_algo=='ss2') ? '3' : ((row.item.smartscan_algo=='ss3') ? '5' : ((row.item.smartscan_algo=='ss4') ? '10' : '2') )}}. More discussion available internally here.
\n\nThis shows the combined allocated size in GB of the SGA and PGA. For Oracle RAC clusters, the sizes are the aggregate across all instances. Use this chart to help select a target memory configuration for this database (be sure to allow for some additional memory for the operating system). This is a stacked chart.
\nThe below chart is the breakout of the combined SGA+PGA by child database. This chart can show you how the workload is spread across the different child databases being consolidated together. This is a stacked chart.
\nDisclaimer: This is not a way to ask support-type questions. This should be used for sizing and architectural types of questions. This Ask a Database Advisor service is a new informal experimental offering and the individuals involved will make a best-effort to respond, so please be patient and understanding.
\nContext for this question (details will be shared with the AWS Database Advisor to help them answer your question):
\n\n[AWS White Paper] Database Caching Strategies
\n\n[AWS Blog] Optimize cost and boost performance of RDS with ElastiCache
\n \n[Webcast Replay] Optimize RDS Costs with ElastiCache
\n \n[Customer Blog - Wiz] Using ElastiCache to improve performance and reduce costs
\n \n \n\n[AWS White Paper] Running Oracle Hypervisors on EC2 Bare Metal
\n \n[NetApp White Paper] Oracle Database and EC2 and FSx-OnTap Best Practices
\n\n[AWS Documentation] FSx-OnTap Performance
\n \n \n \n \n\n{{ ((row.item.IS_PDB=='YES')) ? \"This shows the average # of CPU threads used by the PDB during each snapshot time period using the Oracle metric: (CPU Usage Per Sec from DBA_HIST_CON_SYSMETRIC_SUMM) along with the alternate metric: (CPU_CONSUMED_TIME from DBA_HIST_RSRC_PDB_METRIC). Values at the PDB-level are sanity-checked against the CDB (if found in DBCSI) or against typical thresholds, and outliers are not shown. The value can be greater than 1 when multiple CPU threads at active at the same time. For a RAC cluster, this chart aggregates the usage across all instances. Use this chart to help identify a target baseline of CPU threads needed for this database.\" : \"This shows the average and peak # of CPU threads used by the database during each snapshot time period using the Oracle metric: (CPU Usage Per Sec). The value can be greater than 1 when multiple CPU threads at active at the same time. For a RAC cluster, this chart aggregates the usage across all instances. Use this chart to help identify a target baseline of CPU threads needed for this database.\" }}
\n\n 0)\">The following chart has been adjusted for AIX CPU Usage timekeeping by using an AIX Power SMT CPU Time Adjustment multiplier of {{row.item.AIX_CPU_ADJUSTMENT_USED}}. \n The Average and Peak CPU Usage lines have been adjusted from the source system's threads ({{ row.item.PLATFORM_NAME }}) to x86 VCPUs using the CPU Thread Ratio of {{row.item.cpu_thread_ratio}}.\n{{ row.item.ExadataVCPUAdjustmentText }}
\n[TargetWithCaching is the revised target value after applying your caching assumptions]
\nAnd the below chart is the breakout of the Average DB CPU Usage by RAC Instance. This chart can show you how the workload is spread across the different RAC nodes.
\nAnd the below chart is the breakout of the Average DB CPU Usage by child database. This chart can show you how the workload is spread across the different child databases being consolidated together. This is a stacked chart.
\n{{ ((row.item.IS_PDB=='YES')) ? \"This shows the average IO throughput rate (MB/s) during each snapshot time period using the Oracle metrics: (Physical Read Total Bytes Per Sec, Physical Write Total Bytes Per Sec from DBA_HIST_CON_SYSMETRIC_SUMM) along with the alternate metric: (IOMBPS from DBA_HIST_RSRC_PDB_METRIC). Values at the PDB-level are sanity-checked against the CDB (if found in DBCSI) or against typical thresholds, and outliers are not shown. For a RAC Cluster, the chart aggregates the metrics across all the instances. For an Exadata system, this does include I/O from the storage cells even I/O avoided via storage index use. Use this chart to help select a target baseline for Throughput for this database.\" : \"This shows the average and peak IO throughput rate (MB/s) during each snapshot time period using the Oracle metrics: (Physical Read Total Bytes Per Sec, Physical Write Total Bytes Per Sec). For a RAC Cluster, the chart aggregates the metrics across all the instances. For an Exadata system, this does include I/O from the storage cells even I/O avoided via storage index use. Use this chart to help select a target baseline for Throughput for this database.\" }}
\nThe below chart is the breakout of the Average Throughput by child database. This chart can show you how the workload is spread across the different child databases being consolidated together. This is a stacked chart.
\nThe below chart shows the estimated I/O Throughput after reflecting post-Exadata tuning activities {{row.item.ExadataAdjustmentText}}.
\nThe tuning logic is to start with the target value for I/O throughput as seen on the source Exadata ({{row.item.OrigThroughput}} MB/s). The {{(row.item.smartscan_algo=='ss2') ? 'MODERATE' : ((row.item.smartscan_algo=='ss3') ? 'HEAVY' : ((row.item.smartscan_algo=='ss4') ? 'VERY HEAVY' : 'LIGHT'))}} tuning assumption assumes that I/O Throughput can be reduced by a factor of {{(row.item.smartscan_algo=='ss2') ? '3' : ((row.item.smartscan_algo=='ss3') ? '5' : ((row.item.smartscan_algo=='ss4') ? '10' : '2') )}}. More discussion available internally here.
\n \nThe below chart shows the average IOPS during each snapshot time period using the Oracle metric: (IOPS from DBA_HIST_RSRC_PDB_METRIC). This is a PDB-specific metric and may not include some IOPS that happen at the CDB-level. For a RAC Cluster, the chart aggregates the metrics across all the instances. Use this chart to help select a target baseline for IOPS for this database.
\nThe below chart shows the average and peak rate (Database IO Requests/second) during each snapshot time period using the Oracle metrics: (Physical Read Total IO Requests Per Sec, Physical Write Total IO Requests Per Sec). For a RAC Cluster, the chart aggregates the metrics across all the instances. Use this chart to help select a target baseline for IOPS for this database.
\nThe below chart is the breakout of the Average IOPS by child database. This chart can show you how the workload is spread across the different child databases being consolidated together. This is a stacked chart.
\nThe below chart shows the estimated IOPS after reflecting post-Exadata tuning activities {{row.item.ExadataAdjustmentText}}.
\nThe tuning logic is to start with the target value for IOPS as seen on the source Exadata ({{row.item.OrigIOPS}}). The {{(row.item.smartscan_algo=='ss2') ? 'MODERATE' : ((row.item.smartscan_algo=='ss3') ? 'HEAVY' : ((row.item.smartscan_algo=='ss4') ? 'VERY HEAVY' : 'LIGHT'))}} tuning assumption assumes that IOPS can be reduced by a factor of {{(row.item.smartscan_algo=='ss2') ? '3' : ((row.item.smartscan_algo=='ss3') ? '5' : ((row.item.smartscan_algo=='ss4') ? '10' : '2') )}}. More discussion available internally here.
\n\nThis shows the combined allocated size in GB of the SGA and PGA. For Oracle RAC clusters, the sizes are the aggregate across all instances. Use this chart to help select a target memory configuration for this database (be sure to allow for some additional memory for the operating system). This is a stacked chart.
\n[TargetWithCaching is the revised target value after applying your caching assumptions]
\nThe below chart is the breakout of the combined SGA+PGA by child database. This chart can show you how the workload is spread across the different child databases being consolidated together. This is a stacked chart.
\nThis feature lets you share your DBCSI databases via generated obscure URLs, similar to how you share in the AWS Pricing Calculator. The other users who wish to use your generated sharing URL will need to be DBCSI users as well. The generated URL does not copy over your algorithm choices or any manually entered target values. To continue, please read and acknowledge the following:
\nSharing Acknowledgement: Any DBCSI user who has been given the generated sharing URL will be able to see the DBCSI information for these databases.
\n \n\nThis feature lets you download DBCSI data into a CSV. There are multiple formats to choose from. The DETAIL format is the primary choice and is also the right choice for uploading into the Directional Business Case (DBC) tool. The ALTERNATE-1 format is a reduced set of columns in the format used by the AWS-internal ODBCC tool.
\nThis feature lets you create a new Batch containing the {{this.countSelected}} databases you selected.
\nAre you sure you want to delete this batch?
\nAre you sure you want to delete this batch AND its databases?
\n