Findings

Click Analysis link to see the detail of findings. You can find the log files here. If collected, you can find the System Activity Report (SAR) files under the guestosfiles folder.

Category Analysis High Medium Low
No issues found.

Run Information

VMName
ReportId
OS Type
PerfInsights Version
CommandLine
Scenario
SR Number
Start Time (UTC)
End Time (UTC)
Logs Log files folder

Machine Information

VM Name
VM Size
VM Location
OS Type
Azure Environment
Subscription Id
Resource Group
Vm Id
Publisher
Offer
Image Sku
Image Version
Number of Cores
Number of Logical Processors
Threads Per Core
Availability Zone
Virtual Machine Scale Set Name
Placement Group Id
Tags
Max IOPS
Max Throughput MB/s
Storage Tier
Public IP (Network Card #0)
Private IP (Network Card #0)


No issues found.

No high CPU usage period and consumer found.
Performance Diagnostics Analysis Period:
Notes:
  • Total CPU is the amount of time in milliseconds the process was running on the CPU.
  • All dates and times are shown in UTC.
  • Some usage spikes may be due to diagnostic tools run by the HPC scenario.
  • Performance counter data was collected and is available under the output folder.

High CPU Usage Periods


StartTime EndTime All Processors CPU Usage(ms) All Processors CPU Usage(%) Processors High CPU Usage Breakdown
Processor Number CPU Usage(ms) CPU Usage(%)

Top long running CPU consumers

The graph below shows the periods where the top long running CPU consumers were busy and their CPU usage data.

The table below lists the top long running CPU consumers during Performance Diagnostics analysis period. Sorted by Total CPU Usage(ms) in descending order.
Process Total CPU Usage(ms) Total CPU Usage(%) Processes High CPU Usage Breakdown Parent Process
Child Processes
CPU Usage(ms) Start Time End Time

Top spike CPU consumers

The graph below shows the periods where the top spike CPU consumers were busy and their CPU usage data. Spike CPU consumers are the processes contributing to high Total CPU Usage % during the monitoring period.

The table below lists the top spike CPU consumers during Performance Diagnostics analysis period. Sorted by Total CPU Usage(%) in descending order.
Process Total CPU Usage(ms) Total CPU Usage(%) Processes High CPU Usage Breakdown
CPU Usage(ms) Start Time End Time


No issues found.


No issues found.

System Memory Summary

Total Memory (MB) Used Memory (MB) Free Memory (MB) Shared Memory (MB) Buffers Used (MB) Cache (MB) Available Memory (MB) Used Memory % Free Memory % Available Memory %

Swap Usage

Name Type Size (MB) Used (MB) Priority Used %
No swap space found.

No high memory usage period and consumer found.
Performance Diagnostics Analysis Period:
Notes:
  • All dates and times are shown in UTC.
  • Some usage spikes may be due to diagnostic tools run by the HPC scenario.

High Memory Usage Periods


StartTime EndTime AvailablePhysicalMemory%

Top Memory consumers

The graph below shows the periods where the top long running memory consumers had high memory usage.

The table below lists the top memory consumers during Performance Diagnostics analysis period. Sorted by AvgResidentMemoryKB in descending order.
Process AvgResidentMemory AvgSwapMemory PeakResidentMemory PeakSwapMemory ProcessesHighMemoryUsageBreakdown
AvgResidentMemory AvgSwapMemory PeakResidentMemory PeakSwapMemory StartTime EndTime


No issues found.

Block Devices

No block device data found.
Block Device IsOSDisk IsTempDisk IsDataDisk IsLVMDevice IsMDADMDevice IsLVMMember IsMDADMMember Size (GB) SCSIBus SCSIPort SCSITargetId SCSILun

Partitions

No partition data found.
Partition Name File System Size (GB) Available Space (GB) IsOSDisk IsTempDisk IsDataDisk

Mount Points

No mount point data found.
Mount Point Device Name File System Size (GB) Available Space (GB) IsOSDisk IsTempDisk IsDataDisk IsLVM IsMDADM IsStriped IsSpanned IsRaid5

LVM

No LVM data found.
Volume Group Format Size (GB) UUID Physical Extent Size Member Devices Count
Member Devices
Logical Volume Count
Logical Volume Properties
Name
LVM Block Device
Path
Size (GB)
Mapper
Mount Point
RAID Type
Segments Number
UUID


MDADM

No MDADM data found.
Block Device Array Size (GB) Chunk Size (KB) Creation Time Version Status Raid Total Devices
Member Devices
Active Devices Failed Devices Spare Devices Working Devices RAID Devices
Partitions
Mount Points

No high disk usage period and consumer found.
Performance Diagnostics Analysis Period:
Notes:
  • All dates and times are shown in UTC.
  • Some usage spikes may be due to diagnostic tools run by the HPC scenario.

High Disk Usage Periods

Start Time End Time Disks High Usage Breakdown
Device Sector Size (Bytes) Avg IOPS Avg Throughput (KBytes/s) Avg Throughput (MBytes/s) Read Count Kilobytes Read Write Count Kilobytes Write Max Queue Depth

Top long running Disk consumers

The table below lists the top long running disk consumers during Performance Diagnostics analysis period. Sorted by TotalDiskIoCount in descending order.
Process TotalDiskIoCount TotalDiskIoSize AvgIOPS ProcessesHighDiskUsageBreakdown
ReadCount BytesRead WriteCount BytesWritten StartTime EndTime

Top spike Disk consumers

The table below lists the top spike disk consumers during Performance Diagnostics analysis period. Sorted by AvgDiskIops in descending order.
Process TotalDiskIoCount TotalDiskIoSize AvgIOPS ProcessesHighDiskUsageBreakdown
ReadCount BytesReadKB WriteCount BytesWriteKB StartTime EndTime
No high resolution disk data is available.

Performance Diagnostics Analysis Period:
Notes:
  • All dates and times are shown in UTC.
  • Click timeline chart to make it interactive and scroll mouse wheel to zoom in or out.
Historical High Resolution Disk Data:
  • Historical data files are located under output folder.
  • Choose a file to view data in IOPS/Throughput graphs below.
DataFileName MinDataTime MaxDataTime CountOfRecords

Physical Disk IOPS Graph

The graph below shows IOPS per physical disk during the analysis period. The second Y-Axis shows current queue length.
Azure Storage IOPS Throttling Limit
IOPS and Queue Depth

Physical Disk Throughput Graph

The graph below shows Throughput(MBPS) per physical disk during the analysis period. The second Y-Axis shows current queue length.
Azure Storage Throughput(MBPS) Throttling Limit
Throughput(MBPS) and Queue Depth



No issues found.

General Network Configuration

No information found.

Interface Summary

Alias NlMtu IPv4 IPv6 Mask MAC Network Acceleration Enabled

Interface Statistics

Alias collisions multicast rx_bytes rx_compressed rx_crc_errors rx_dropped rx_errors rx_fifo_errors rx_frame_errors rx_length_errors rx_missed_errors rx_nohandler rx_over_errors rx_packets tx_aborted_errors tx_bytes tx_carrier_errors tx_compressed tx_dropped tx_errors tx_fifo_errors tx_heartbeat_errors tx_packets tx_window_errors

Routing Tables

Destination Flags Gateway IRTT Iface MTU Mask Metric RefCnt Use Window

Netstat

No information found.
Number of connections grouped by State
  • LISTEN:
  • ESTABLISHED:
  • TIME_WAIT:
  • Stateless:

Protocol State Local Address Local Port Remote Address Remote Port Process Name PID User


No data found.

Guest OS Information

Name
Pretty Name
Version
Version ID
Current Kernel
Time Zone
Temp Disk Mountpoint
Swap File Location

Running Processes

No running process data found.
PID Process Name State CPU Usage (ms) Resident Memory

Guest Agent Information

Binary Path
Version

Drivers Information

Driver Name Number of Instances Dependencies Size In Bytes State

LIS (Linux Integration Services) Drivers Information

Name Description

Kernel Information

No information found.
Bug Report URL
Home URL
Id
Id like
Name
Pretty Name
Support URL
Ubuntu Codename
Version
Version Codename
Version ID
Current Kernel
Available Kernels
Kernel Configurations
Kernel Initrd Images
Kernel System Map

PCI Information

No information found.
Device Name Vendor

CPU Information

No information found. This machine runs with logical processors.

CPU #

Address Sizes
Apicid
Bogomips
Bugs
Cache Alignment
Cache Size
CL Flush Size
Core ID
MHz
Cores
Family
CPU ID Level
Flags
fpu
fpu_exception
initialapicid
microcode
model
modelname
physicalid
powermanagement
processor
siblings
stepping
vendor_id
wp
BogoMips
CPU Architecture
CPU Implementer
CPU Part
CPU Revision
CPU Variant
Features
Processor

GPU Information


No GPU information found. This machine runs with GPUs.

GPU #

Id
Model
Driver Version
Serial Number
UUID
VBIOS Version


What is PerfInsights?
PerfInsights is a self-help diagnostics tool that collects and analyzes the diagnostic data, and provides a report to help troubleshoot Linux virtual machine performance problems in Azure. PerfInsights Linux documentation is here.

Where do I start to review this report?
Start with the Overview tab. Review the recommendations and links for the findings. Learn about how they can affect performance, and also about best practices for performance-optimized configurations. Learn more about reviewing the report here.

What kind of data is collected by PerfInsights?
PerfInsights collects information about the Linux VM, operating system, block devices, high resource consumers, configuration, and various logs. Details about what kind of information is collected by PerfInsights is here.