描述
此表包含可即时计算并以 Prometheus 格式导出的维度指标,且会始终保持最新状态。
列
metric(String) — 指标名称。value(Float64) — 指标值。description(String) — 指标说明。labels(Map(String, String)) — 指标标记。
别名:
name—metric的别名。
示例
您可以使用如下查询,以 Prometheus 格式导出所有维度指标。
SELECT
metric AS name,
toFloat64(value) AS value,
description AS help,
labels,
'gauge' AS type
FROM system.dimensional_metrics
FORMAT Prometheus指标说明
merge_failures
自启动以来所有失败 merge 的数量。
startup_scripts_failure_reason
按错误类型指示启动脚本失败情况。当启动脚本失败时,该值设为 1,并带有错误名称标记。
merge_tree_parts
MergeTree 数据分区片段的数量,按分区片段状态、分区片段类型以及是否为投影分区片段区分。
filesystem_cache_evictions_total
从文件系统缓存中逐出的 File 段数量,按缓存名称标记。默认禁用;可通过 expose_prometheus_eviction_metrics 启用。
filesystem_cache_evicted_bytes_total
从文件系统缓存中被逐出的 File 段总字节数,按缓存名称加标签。默认禁用;可通过 expose_prometheus_eviction_metrics 启用。
filesystem_cache_evictions_by_user_total
从文件系统缓存中逐出的 File 段数量,按缓存名称和用户 ID 标记。默认禁用;可通过 expose_prometheus_eviction_metrics 和 expose_prometheus_eviction_metrics_per_user 启用。
filesystem_cache_evicted_bytes_by_user_total
从文件系统缓存中淘汰的 File 段总字节数,按缓存名称和用户 ID 进行标记。默认禁用;可通过 expose_prometheus_eviction_metrics 和 expose_prometheus_eviction_metrics_per_user 启用。
object_storage_queue_failures_total
按数据库、表、处理阶段 (read、set_processing、insert、commit) 和错误代码标记的 ObjectStorageQueue (S3Queue/AzureQueue) 失败次数。
object_storage_queue_permanently_failed_files_total
ObjectStorageQueue (S3Queue/AzureQueue) 在耗尽重试次数后 (或禁用重试时) 永久放弃处理的文件数量,按数据库和表标记。每个此类文件的数据都将永远不会被处理。
object_storage_queue_newest_seen_object_timestamp_seconds
ObjectStorageQueue (S3Queue/AzureQueue) 表截至目前已发现的最新对象的最后修改时间的 Unix 时间戳,按数据库和表标记。
object_storage_queue_newest_committed_object_timestamp_seconds
截至目前由 ObjectStorageQueue (S3Queue/AzureQueue) 表完全处理的最新对象的最后修改时间的 Unix 时间戳,按数据库和表标记。
另请参阅
- system.asynchronous_metrics — 包含定期计算的指标。
- system.events — 包含已发生的一系列事件。
- system.metric_log — 包含来自表
system.metrics和system.events的指标值历史记录。 - 监控 — ClickHouse 监控的基本概念。