Tencent Cloud CAT¶
Collect Tencent Cloud CAT task objects and probe metrics.
Configuration¶
Install Func¶
We recommend enabling the hosted Func edition under Guance Integration - Extensions. The prerequisites are installed automatically.
If you deploy Func yourself, see Self-hosted Func deployment.
Install Scripts¶
Prerequisites
- Create an access key: In Tencent Cloud
Cloud Access Management, create an API key for the sub-user used for collection, and save theSecretIdandSecretKey. - Configure permissions: Grant the global preset policy
ReadOnlyAccessto the CAM sub-user that owns the API key. If you use a custom least-privilege policy, allowcat:DescribeProbeTasksandcat:DescribeProbeMetricData.
Install the following collectors from the DataFlux Func Script Market:
Integration (TencentCloud-CAT Collector), script set IDintegration_tencentcloud_cat. It collects CAT task objects and can start task-object and metric collection.Integration (TencentCloud-CAT Metric Collector), script set IDintegration_tencentcloud_cat_metric. It collects CAT metrics.
After clicking Install, enter the following parameters:
TencentCloud SecretId: Tencent Cloud access key ID. Required.TencentCloud SecretKey: Tencent Cloud access key secret. Required.Your Account Name: Custom account name. Optional.
CAT tasks are not region-specific, so the collector does not require region configuration.
Click Deploy Startup Script. The system automatically creates the Startup script set and configures the corresponding startup script.
After enabling it, view the corresponding automatic trigger configuration under Management / Automatic Trigger Configuration. Click Run to execute it immediately without waiting for the schedule. After a short wait, you can review the execution records and logs.
Verification¶
- Under Management / Automatic Trigger Configuration, confirm that the task has an automatic trigger configuration, and check its execution records and logs for errors.
- View object data under Guance Infrastructure / Custom:
tencentcloud_cat_task. - View the corresponding measurements under Guance Metrics:
tencentcloud_cat_browse,tencentcloud_cat_network,tencentcloud_cat_transport,tencentcloud_cat_upload_download,tencentcloud_cat_media_stream,tencentcloud_cat_whois.
Metrics¶
CAT writes data to different measurements according to task type:
| Task Type | measurement |
|---|---|
| Page Performance | tencentcloud_cat_browse |
| Network Quality | tencentcloud_cat_network |
| Port Performance | tencentcloud_cat_transport |
| File Transfer | tencentcloud_cat_upload_download |
| Audio/Video Experience | tencentcloud_cat_media_stream |
| Domain Whois | tencentcloud_cat_whois |
Common Tags¶
The following common tags are retained by the metric collector. The tag names shown are those written to Guance and use lowercase snake_case.
| Field | Description |
|---|---|
area |
Probe node region |
area_id |
Probe node region ID |
city |
Probe node city |
city_code |
Probe node city code |
cloud_provider |
Cloud provider identifier, fixed to tencentcloud |
destination_city_name |
Destination city name; may be returned for file-transfer tasks |
distinct |
Probe node region display name |
error_id |
Error ID |
error_info |
Error information |
operator |
Probe node carrier |
operator_id |
Probe node carrier ID |
probe_app_id |
Probe application ID |
probe_nodes_code |
Probe node code |
target_address |
Probe target address |
task_category |
Probe task category |
task_id |
Probe task ID |
task_name |
Probe task name |
task_type |
Probe task type |
task_type_name |
Probe task type name |
type |
Upload/download type; may be returned for file-transfer tasks |
The host value in metric data is not written as a tag to avoid conflicts with the built-in host tag in Guance. Task names consistently use the task_name supplied by the tencentcloud_cat_task object.
Measurement¶
tencentcloud_cat_browse
Raw Metrics¶
| metricName | description | unit |
|---|---|---|
average_dns_query_time |
Average DNS query time | ms |
average_download_time |
Average download time | ms |
average_request_time |
Average request time | μs |
average_response_time |
Average response time | ms |
average_ssl_time |
Average SSL handshake time | ms |
average_tcp_connection_time |
Average TCP connection time | ms |
block_projection_time |
blocking-phase estimated time | N/A |
bytes_received |
receive byte count | B |
count |
Probe Total count | count |
d_time |
Page Overall load time | ms |
dns_error_count |
DNS Error count | count |
dns_look_up |
DNS query time | ms |
dns_query_time |
DNS query time | ms |
document_finish_time |
Document completion time | ms |
dom_load_time |
DOM load time | ms |
element_hijack_count |
element hijack count | count |
err_ele_count |
Error element count | count |
error_count |
Probe Failed count | count |
file_download_size |
download File size | B |
file_download_speed |
File download speed | N/A |
file_download_time |
File download time | ms |
first_ele_all_num |
first-screen element count | count |
first_err_ele_num |
first-screen Error element count | count |
first_full_time |
First full load time | ms |
first_packet_arrival_time |
First packet arrival time | ms |
first_screen_time |
first-screen time | ms |
hijack_302_count |
302 hijack count | count |
hijack_302_success_count |
302 hijack Successful count | count |
hijack_count |
hijack count | count |
hijack_error_count |
hijack Error count | count |
hijack_normal_count |
Normal hijack count | count |
hijack_third_count |
Third-party hijack count | count |
host_hijack_count |
Host hijack count | count |
http_error_count |
HTTP Error count | count |
k100_time |
100 KB download time | ms |
onload_time |
Window onload time | ms |
overall_download_speed |
Overall download speed | KB/s |
page_3xx_num |
Page 3xx redirect count | count |
page_start_render |
Page start rendering time | ms |
projection_dns_query_time |
DNS query estimated time | N/A |
projection_download_time |
download estimated time | N/A |
projection_request_time |
request estimated time | N/A |
projection_response_time |
response estimated time | N/A |
projection_ssl_time |
SSL estimated time | N/A |
projection_tcp_connection_time |
TCP connection estimated time | N/A |
redirect_count |
redirect count | count |
redirect_time |
redirect time | ms |
render_speed |
render speed | N/A |
render_time |
render time | ms |
request_time |
request time | ms |
response_time |
response time | ms |
round_trips |
round trip count | count |
slow_count |
Slow Probe count | count |
ssl_error_count |
SSL Error count | count |
ssl_time |
SSL handshake time | ms |
success_count |
Probe Successful count | count |
tcp_connection_time |
TCP connection time | ms |
tcp_connects |
TCP connection count | count |
tcp_error_count |
TCP Error count | count |
trans_error_count |
transfer Error count | count |
Calculated Metrics¶
The following metrics are calculated from raw metrics within the same probe cycle to show page-performance transfer speed and success rate. The dimensions are task_id, area, city, operator, probe_nodes_code.
| metricName | description | unit |
|---|---|---|
overall_transfer_speed |
Overall transfer speed, Formula: (sum("bytes_received") / 1024) / (sum("d_time") / 1000) |
KB/s |
success_rate |
Probe success rate, Formula: sum("success_count") / sum("count") * 100 |
percent |
Measurement¶
tencentcloud_cat_network
Raw Metrics¶
| metricName | description | unit |
|---|---|---|
count |
Probe Total count | count |
dig_error_count |
DNS resolution Error count | count |
dig_success_count |
DNS resolution Successful count | count |
error_count |
Probe Failed count | count |
hijack_count |
hijack count | count |
ping_count |
Ping packet count | count |
ping_drop_package_count |
Ping packet loss count | count |
ping_drop_package_percent |
Ping packet loss percentage | N/A |
ping_error_count |
Ping Error count | count |
ping_success_count |
Ping Successful count | count |
ping_time |
Ping latency | ms |
ping_timeout_num |
Ping timeout count | count |
query_time |
DNS query time | ms |
success_count |
Probe Successful count | count |
tracert_avg_time |
Average Tracert latency | ms |
tracert_error_count |
Tracert Error count | count |
tracert_skip_count |
Tracert skip count | count |
tracert_success_count |
Tracert Successful count | count |
ttl_time |
TTL value | s |
Calculated Metrics¶
The following metrics are calculated from raw metrics within the same probe cycle to show overall network-probe duration, success rate, DNS resolution success rate, Ping packet loss rate, and more. The dimensions are task_id, area, city, operator, probe_nodes_code.
| metricName | description | unit |
|---|---|---|
dns_error_rate |
DNS resolution error rate, Formula: sum("dig_error_count") / (sum("dig_success_count") + sum("dig_error_count")) * 100 |
percent |
dns_success_rate |
DNS resolution success rate, Formula: sum("dig_success_count") / (sum("dig_success_count") + sum("dig_error_count")) * 100 |
percent |
packet_loss_rate |
Ping packet loss rate, Formula: sum("ping_drop_package_count") / sum("ping_count") * 100 |
percent |
ping_success_rate |
Ping success rate, Formula: sum("ping_success_count") / (sum("ping_success_count") + sum("ping_error_count")) * 100 |
percent |
success_rate |
Probe success rate, Formula: sum("success_count") / (sum("success_count") + sum("error_count")) * 100 |
percent |
total_time |
Network QualityOverall time, Formula: avg("query_time") + avg("ping_time") |
ms |
Measurement¶
tencentcloud_cat_transport
Raw Metrics¶
| metricName | description | unit |
|---|---|---|
avg_error_rate |
Average error rate | N/A |
client_net_speed |
Client network speed | N/A |
connection_time |
connection time | ms |
count |
Probe Total count | count |
d_time |
Overall time | ms |
dns_error_count |
DNS Error count | count |
dns_time |
DNS resolution time | ms |
error_count |
Probe Failed count | count |
max_time |
Maximum time | ms |
min_time |
Minimum time | ms |
request_time |
request time | ms |
response_time |
response time | ms |
send_true_package_num |
Successful send packet count | count |
ssl_error_count |
SSL Error count | count |
ssl_time |
SSL handshake time | ms |
success_count |
Probe Successful count | count |
tcp_error_count |
TCP Error count | count |
trans_error_count |
transfer Error count | count |
verify_error_rate |
verification error rate | N/A |
wait_time |
wait time | ms |
Calculated Metrics¶
The following metrics are calculated from raw metrics within the same probe cycle to show port-probe and verification success rates. The dimensions are task_id, area, city, operator, probe_nodes_code.
| metricName | description | unit |
|---|---|---|
success_rate |
Probe success rate, Formula: sum("success_count") / (sum("success_count") + sum("error_count")) * 100 |
percent |
verify_success_rate |
verification success rate, Formula: (1 - sum("verify_error_rate") / (sum("verify_error_rate") + (sum("success_count") * 100))) * 100 |
percent |
Measurement¶
tencentcloud_cat_upload_download
Raw Metrics¶
| metricName | description | unit |
|---|---|---|
avg_speed |
Average transfer speed | N/A |
count |
Probe Total count | count |
dns_time |
DNS resolution time | ms |
error_count |
Probe Failed count | count |
file_size |
File size | B |
first_package_time |
First packet time | ms |
hijack_count |
hijack count | count |
host_hijack_count |
Host hijack count | count |
rec_time |
receive time | ms |
redirect_count |
redirect count | count |
redirect_times |
redirect count | count |
response_time |
response time | ms |
send_time |
send time | ms |
slow_count |
Slow Probe count | count |
ssl_time |
SSL handshake time | ms |
success_count |
Probe Successful count | count |
tcp_time |
TCP connection time | ms |
total_time |
Total transfer time | ms |
transfer_time |
transfer time | ms |
Calculated Metrics¶
The following metrics are calculated from raw metrics within the same probe cycle to show the success rate of file upload or download tasks. The dimensions are task_id, area, city, operator, probe_nodes_code, type.
| metricName | description | unit |
|---|---|---|
success_rate |
File Transfersuccess rate, Formula: sum("success_count") / sum("count") * 100 |
percent |
Measurement¶
tencentcloud_cat_media_stream
Raw Metrics¶
| metricName | description | unit |
|---|---|---|
audio_rate |
audio bitrate | N/A |
avg_down_speed |
Average download speed | N/A |
buffer_num |
buffering count | count |
caton_flag |
stall flag | N/A |
caton_number |
stall count | count |
caton_rate |
stall rate | N/A |
caton_time |
stall duration | ms |
client_net_speed |
Client network speed | N/A |
connect_time |
connection time | ms |
connecting_time |
connection establishment time | ms |
count |
Probe Total count | count |
dns_resolution_time |
DNS resolution time | ms |
download_size |
download size | B |
error_count |
Probe Failed count | count |
first_buffer_time |
First buffering time | ms |
first_down_size |
First download size | B |
first_down_speed |
First download speed | N/A |
first_duration_time |
First playback duration | N/A |
first_frame |
First frame time | ms |
first_plag_time |
First playback stall time | N/A |
fps |
frame rate | N/A |
hijack_count |
hijack count | count |
hijack_domain_count |
Domain hijack count | count |
page_first_time |
Page first-byte time | ms |
page_one_byte_time |
Page single-byte time | N/A |
resource_ssl |
Resource SSL time | N/A |
response_time |
response time | ms |
slow_count |
Slow Probe count | count |
success_count |
Probe Successful count | count |
total_buffer_time |
Total buffering duration | ms |
total_down_time |
Total download time | ms |
total_plag_time |
Total playback stall time | N/A |
video_rate |
video bitrate | N/A |
wait_rate |
wait rate | N/A |
Calculated Metrics¶
The following metrics are calculated from raw metrics within the same probe cycle to show the success rate of audio/video probes. The dimensions are task_id, area, city, operator, probe_nodes_code.
| metricName | description | unit |
|---|---|---|
success_rate |
audio/video Probe success rate, Formula: sum("success_count") / (sum("success_count") + sum("error_count")) * 100 |
percent |
Tip
The metric-point interval matches the probe frequency configured for the Tencent Cloud CAT task.