gcp-logs-monitoring

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Goal

目标

Inspect Cloud Logging and Cloud Monitoring data quickly and repeatably from the terminal.

从终端快速且可重复地检查Cloud Logging和Cloud Monitoring数据。

Inputs to collect (ask only if missing)

需要收集的输入信息（仅在缺失时询问）

```
project_id
```
```
time_window
```
for investigation (for example: last
```
30m
```
,
```
2h
```
, or explicit UTC
```
start/end
```
)

service/resource context

(

cloud_run_revision

k8s_container

gce_instance

, load balancer, etc.)

```
signals of interest
```
(errors, latency, CPU, memory, request count, restarts)
```
output_format
```
(
```
table
```
for quick scans,
```
json
```
for deeper analysis)

```
project_id
```
调查的
```
time_window
```
（例如：最近
```
30m
```
、
```
2h
```
，或明确的UTC
```
start/end
```
时间）

服务/资源上下文

（

cloud_run_revision

、

k8s_container

、

gce_instance

、负载均衡器等）

```
关注的信号
```
（错误、延迟、CPU、内存、请求数、重启次数）
```
输出格式
```
（
```
table
```
用于快速扫描，
```
json
```
用于深度分析）

Execution workflow

执行流程

Validate prerequisites:

bash .agents/skills/gcp-logs-monitoring/scripts/check_prereqs.sh --project <project_id>

Query Cloud Logging:

bash .agents/skills/gcp-logs-monitoring/scripts/read_logs.sh --project <project_id> --filter '<LOG_FILTER>' --freshness 1h --limit 100 --format json

Query Cloud Monitoring time series:

bash .agents/skills/gcp-logs-monitoring/scripts/read_metrics.sh --project <project_id> --filter '<METRIC_FILTER>' --start <UTC_ISO8601> --end <UTC_ISO8601> --format json

Correlate timestamps between logs and metrics, then summarize likely root cause and next checks.

验证前置条件：

bash .agents/skills/gcp-logs-monitoring/scripts/check_prereqs.sh --project <project_id>

查询Cloud Logging：

bash .agents/skills/gcp-logs-monitoring/scripts/read_logs.sh --project <project_id> --filter '<LOG_FILTER>' --freshness 1h --limit 100 --format json

查询Cloud Monitoring时间序列：

bash .agents/skills/gcp-logs-monitoring/scripts/read_metrics.sh --project <project_id> --filter '<METRIC_FILTER>' --start <UTC_ISO8601> --end <UTC_ISO8601> --format json

关联日志和指标之间的时间戳，然后总结可能的根本原因和后续检查建议。

Common filter templates

常用过滤模板

Cloud Logging

Cloud Run errors:

resource.type="cloud_run_revision" severity>=ERROR

GKE container errors:

resource.type="k8s_container" severity>=ERROR

HTTP 5xx in load balancer logs:

resource.type="http_load_balancer" jsonPayload.statusDetails=~"5.."

Timeout text search:

textPayload:"timeout" OR jsonPayload.message:"timeout"

Cloud Run错误：

resource.type="cloud_run_revision" severity>=ERROR

GKE容器错误：

resource.type="k8s_container" severity>=ERROR

负载均衡器日志中的HTTP 5xx错误：

resource.type="http_load_balancer" jsonPayload.statusDetails=~"5.."

超时文本搜索：

textPayload:"timeout" OR jsonPayload.message:"timeout"

Cloud Monitoring

Cloud Run request count:

metric.type="run.googleapis.com/request_count" AND resource.type="cloud_run_revision"

Cloud Run request latencies:

metric.type="run.googleapis.com/request_latencies" AND resource.type="cloud_run_revision"

VM CPU utilization:

metric.type="compute.googleapis.com/instance/cpu/utilization" AND resource.type="gce_instance"

Cloud Run请求数：

metric.type="run.googleapis.com/request_count" AND resource.type="cloud_run_revision"

Cloud Run请求延迟：

metric.type="run.googleapis.com/request_latencies" AND resource.type="cloud_run_revision"

VM CPU利用率：

metric.type="compute.googleapis.com/instance/cpu/utilization" AND resource.type="gce_instance"

Guardrails

注意事项

Prefer passing
```
--project
```
on every command instead of changing global gcloud config.
Start with short windows (
```
15m
```
to
```
2h
```
) and widen only when needed.
Use
```
--format json
```
when output will be parsed or compared across sources.
If auth or project checks fail, fix environment first and then re-run queries.

优先在每个命令中传递
```
--project
```
参数，而非修改全局gcloud配置。
从较短的时间窗口开始（
```
15m
```
到
```
2h
```
），仅在需要时扩大范围。
当输出需要被解析或跨源比较时，使用
```
--format json
```
。
如果认证或项目检查失败，先修复环境问题，再重新运行查询。