腾讯云可观测平台 向量存储桶 COS Vectors 监控指标

这篇说说腾讯云向量存储桶(QCE/COSVECTORS)的监控指标与查询参数,涵盖容量、向量数、请求数、时延、QPS等指标,并说明按appid、bucket、index维度入参查询的方法。 – 指标分桶级与索引级两类,统计周期含60s至86400s。 – 查询需指定Namespace及appid、bucket、index三个维度。

相关产品
腾讯云可观测平台 产品页,含版本规格、价格与购买入口。

一、命名空间

Namespace = QCE/COSVECTORS

二、监控指标

指标名 指标中文名 说明 单位 维度 统计规则
BucketCapacityUsed 桶已用容量 对应 bucket 已用容量 Bytes appid
bucket
[ 60s, sum ]
[ 300s, last ]
[ 3600s, last ]
[ 86400s, last ]
BucketVectorCount 桶内向量总数 对应 bucket 向量总数 Count appid
bucket
[ 60s, sum ]
[ 300s, last ]
[ 3600s, last ]
[ 86400s, last ]
IndexCapacityPut 索引写入数据量 对应 index 写入数据量 Bytes appid
bucket
index
[ 60s, sum ]
[ 300s, sum ]
[ 3600s, sum ]
[ 86400s, sum ]
IndexCapacityQuery 索引检索数据量 对应 index 检索数据量 Bytes appid
bucket
index
[ 60s, expr ]
[ 300s, sum ]
[ 3600s, sum ]
[ 86400s, sum ]
IndexCapacityUsed 索引已用容量 对应 index 已用容量 Bytes appid
bucket
index
[ 60s, last ]
[ 300s, last ]
[ 3600s, last ]
[ 86400s, last ]
IndexRequestCountOther 索引其他请求数 对应 index 的 get、delete、list 操作请求数 Count appid
bucket
index
[ 60s, expr ]
[ 300s, sum ]
[ 3600s, sum ]
[ 86400s, sum ]
IndexRequestCountPut 索引写入请求数 对应 index 写入向量请求数 Count appid
bucket
index
[ 60s, sum ]
[ 300s, sum ]
[ 3600s, sum ]
[ 86400s, sum ]
IndexRequestCountQuery 索引检索请求数 对应 index 检索请求数 Count appid
bucket
index
[ 60s, sum ]
[ 300s, sum ]
[ 3600s, sum ]
[ 86400s, sum ]
IndexRequestCountQuerySuccess 索引检索成功次数 对应 index 检索成功次数 Count appid
bucket
index
[ 60s, sum ]
[ 300s, sum ]
[ 3600s, sum ]
[ 86400s, sum ]
IndexRequestCountSuccess 索引成功请求数 对应 index 成功请求数 Count appid
bucket
index
[ 60s, sum ]
[ 300s, sum ]
[ 3600s, sum ]
[ 86400s, sum ]
IndexRequestCountTotal 索引请求总数 对应 index 请求总数 Count appid
bucket
index
[ 60s, sum ]
[ 300s, sum ]
[ 3600s, sum ]
[ 86400s, sum ]
IndexRequestLatencyAvg 索引平均请求时延 对应 index 平均请求时延 ms appid
bucket
index
[ 60s, avg ]
[ 300s, avg ]
[ 3600s, avg ]
[ 86400s, avg ]
IndexRequestLatencyP95 索引 p95请求时延 对应 index 的 p95请求时延 ms appid
bucket
index
[ 60s, avg ]
[ 300s, avg ]
[ 3600s, avg ]
[ 86400s, avg ]
IndexRequestLatencyP99 索引 p99请求时延 对应 index 的 p99请求时延 ms appid
bucket
index
[ 60s, avg ]
[ 300s, avg ]
[ 3600s, avg ]
[ 86400s, avg ]
IndexRequestLatencyQueryAvg 索引向量检索平均请求时延 对应 index 向量检索平均请求时延 ms appid
bucket
index
[ 60s, avg ]
[ 300s, avg ]
[ 3600s, avg ]
[ 86400s, avg ]
IndexRequestLatencyQueryP95 索引 p95检索时延 对应 index 的 p95检索时延 ms appid
bucket
index
[ 60s, avg ]
[ 300s, avg ]
[ 3600s, avg ]
[ 86400s, avg ]
IndexRequestLatencyQueryP99 索引 p99检索时延 对应 index 的 p99检索时延 ms appid
bucket
index
[ 60s, avg ]
[ 300s, avg ]
[ 3600s, avg ]
[ 86400s, avg ]
IndexRequestQps 索引请求 QPS 对应 index 请求 QPS Count/s appid
bucket
index
[ 60s, avg ]
[ 300s, avg ]
[ 3600s, avg ]
[ 86400s, avg ]
IndexRequestQpsPut 索引向量写入请求 QPS 对应 index 向量写入请求 QPS Count/s appid
bucket
index
[ 60s, avg ]
[ 300s, avg ]
[ 3600s, avg ]
[ 86400s, avg ]
IndexRequestQpsQuery 索引检索向量请求 QPS 对应 index 向量检索请求 QPS Count/s appid
bucket
index
[ 60s, avg ]
[ 300s, avg ]
[ 3600s, avg ]
[ 86400s, avg ]
IndexVectorCount 索引向量总数 对应 index 向量总数 Count appid
bucket
index
[ 60s, last ]
[ 300s, last ]
[ 3600s, last ]
[ 86400s, last ]

三、各维度对应参数总览

参数名称 维度名称 维度解释 格式
Instances.N.Dimensions.0.Name appid 主账号 appid 的维度名称 输入 String 类型维度名称:appid
Instances.N.Dimensions.0.Value appid 主账号 appid 的维度名称 输入具体 APPID,例如:10001234567
Instances.N.Dimensions.1.Name bucket 向量存储桶的维度名称 输入 String 类型维度名称:bucket
Instances.N.Dimensions.1.Value bucket 向量存储桶的维度名称 输入带 APPID 后缀的桶名,例如:example-bucket-1250000000
Instances.N.Dimensions.2.Name index 向量索引的维度名称 输入 String 类型维度名称:index
Instances.N.Dimensions.2.Value index 向量索引的维度名称 输入索引名称,例如:idx-dim3

四、入参说明

查询向量存储桶 COS Vectors 监控数据,入参取值如下:

&Namespace =QCE/COSVECTORS

&Instances.N.Dimensions.0.Name = appid

&Instances.N.Dimensions.0.Value = 主账号的 APPID

&Instances.N.Dimensions.1.Name = bucket

&Instances.N.Dimensions.1.Value = 向量存储桶名称

&Instances.N.Dimensions.2.Name = index
&Instances.N.Dimensions.2.Value = 向量索引名称

赞 (0)
上一篇 5天前
下一篇 5天前