mon监控埋点指标

╰+哭是因爲堅強的太久メ 2023-02-17 05:05 8阅读 0赞

# 1. perf dump #

## 1.1 cluster ##

<table> 
 <thead> 
  <tr> 
   <th>监控类型</th> 
   <th>监控项</th> 
   <th>说明</th> 
   <th>级别</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>perf dump cluster</td> 
   <td>ceph.cluster.num_mon</td> 
   <td>mon数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_mon_quorum</td> 
   <td>法定mon数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_osd</td> 
   <td>集群osd节点数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_osd_up</td> 
   <td>up状态的osd节点数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_osd_in</td> 
   <td>in状态的osd节点数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>osd_epoch</td> 
   <td>osd的epoch号</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster_total_objects</td> 
   <td>集群objects总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_pg</td> 
   <td>集群pg总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_pg_peering</td> 
   <td>长时间存在peering的pg</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_pool</td> 
   <td>集群pool数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster_total_objects</td> 
   <td>集群objects数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_mon</td> 
   <td>集群mon节点数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_mds_up</td> 
   <td>up状态mds数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_mds_in</td> 
   <td>in状态mds数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.num_mds_failed</td> 
   <td>failed状态mds数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.osd_bytes</td> 
   <td>osd总共大小</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.osd_bytes_used</td> 
   <td>用户占用大小</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.cluster.osd_bytes_avail</td> 
   <td>osd可用的大小</td> 
   <td>&nbsp;</td> 
  </tr> 
 </tbody> 
</table>

## 1.2 leveldb ##

<table> 
 <thead> 
  <tr> 
   <th>监控类型</th> 
   <th>监控项</th> 
   <th>说明</th> 
   <th>级别</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>perf dump leveldb</td> 
   <td>ceph.leveldb.leveldb_get</td> 
   <td>获取的数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_transaction</td> 
   <td>处理的数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_get_latency.avgcount</td> 
   <td>获取延迟队列里面的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_get_latency.sum</td> 
   <td>获取延迟队列里面的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_submit_latency.avgcount</td> 
   <td>提交延迟队列里面的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_submit_latency.sum</td> 
   <td>提交延迟队列里面的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_submit_latency.avgcount</td> 
   <td>提交延迟队列里面的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_submit_latency.sum</td> 
   <td>提交延迟队列里面的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_submit_sync_latency.avgcount</td> 
   <td>提交同步延迟队列里面的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_submit_sync_latency.sum</td> 
   <td>提交同步延迟队列里面的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_compact</td> 
   <td>压缩</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_compact_range</td> 
   <td>压缩范围</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_compact_queue_merge</td> 
   <td>压缩合并队列</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.leveldb.leveldb_compact_queue_len</td> 
   <td>压缩队列长度</td> 
   <td>&nbsp;</td> 
  </tr> 
 </tbody> 
</table>

## 1.3 mon ##

<table> 
 <thead> 
  <tr> 
   <th>监控类型</th> 
   <th>监控项</th> 
   <th>说明</th> 
   <th>级别</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>perf dump mon</td> 
   <td>ceph.mon.num_sessions</td> 
   <td>当前打开的监视器会话数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.mon.sessions_add</td> 
   <td>创建监视器会话数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.mon.sessions_rm</td> 
   <td>监视器中删除会话调用的次数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.mon.sessions_trim</td> 
   <td>调整监视器会话数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.mon.num_elections</td> 
   <td>选举监察员数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.mon.election_cal</td> 
   <td>由监视器启动的选举数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.mon.election_win</td> 
   <td>选举赢得选举数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.mon.election_lose</td> 
   <td>选举流失的选举数</td> 
   <td>&nbsp;</td> 
  </tr> 
 </tbody> 
</table>

## 1.4 paxos ##

<table> 
 <thead> 
  <tr> 
   <th>监控类型</th> 
   <th>监控项</th> 
   <th>说明</th> 
   <th>级别</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>perf dump paxos</td> 
   <td>ceph.paxos.start_leader</td> 
   <td>启动leader角色</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.start_peon</td> 
   <td>启动peon角色</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.restart</td> 
   <td>重启数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.refresh</td> 
   <td>刷新</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.refresh</td> 
   <td>刷新</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.refresh_latency.avgcount</td> 
   <td>延迟刷新平均数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.refresh_latency.sum</td> 
   <td>延迟刷新总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.begin</td> 
   <td>开始处理</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.begin_keys.avgcount</td> 
   <td>开始处理keys的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.begin_keys.sum</td> 
   <td>开始处理keys的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.begin_bytes.avgcount</td> 
   <td>开始处理bytes的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.begin_bytes.sum</td> 
   <td>开始处理bytes的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.begin_latency.avgcount</td> 
   <td>开始处理延迟的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.begin_latency.sum</td> 
   <td>开始处理延迟的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.commit</td> 
   <td>提交数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.commit_keys.avgcount</td> 
   <td>提交keys的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.commit_keys.sum</td> 
   <td>提交keys的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.commit_latency.avgcount</td> 
   <td>提交延迟的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.commit_latency.sum</td> 
   <td>提交延迟的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.collect</td> 
   <td>收集</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.collect_keys.avgcount</td> 
   <td>收集的keys的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.collect_keys.sum</td> 
   <td>收集的keys的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.collect_bytes.avgcount</td> 
   <td>收集的bytes数的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.collect_bytes.sum</td> 
   <td>收集的bytes数的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.collect_latency.avgcount</td> 
   <td>收集延迟平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.collect_latency.sum</td> 
   <td>收集延迟总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.collect_uncommitted</td> 
   <td>&nbsp;</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.collect_timeout</td> 
   <td>收集超时时间</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.accept_timeout</td> 
   <td>接受超时时间</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.lease_ack_timeout</td> 
   <td>租约确认超时时间</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.lease_timeout</td> 
   <td>租约超时时间</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.store_state</td> 
   <td>存储的状态</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.store_state_keys.avgcount</td> 
   <td>存储状态中的事务密钥平均数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.store_state_keys.sum</td> 
   <td>存储状态中的事务密钥总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.store_state_bytes.avgcount</td> 
   <td>存储状态中事务中的数据的平均数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.store_state_bytes.sum</td> 
   <td>存储状态中事务中的数据的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.store_state_latency.avgcount</td> 
   <td>存储状态延迟平均数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.store_state_latency.sum</td> 
   <td>存储状态延迟总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.share_state</td> 
   <td>共享状态</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.share_state_keys.avgcount</td> 
   <td>共享状态的keys的平均数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.share_state_keys.sum</td> 
   <td>共享状态的keys的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.share_state_bytes.avgcount</td> 
   <td>共享状态数据平均数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.share_state_bytes.sum</td> 
   <td>共享状态数据总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.new_pn</td> 
   <td>新建提议号询问</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.new_pn_latency.avgcount</td> 
   <td>新建提议号询问等待时间的平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>ceph.paxos.new_pn_latency.sum</td> 
   <td>新建提议号询问等待时间的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
 </tbody> 
</table>

## 1.5 throttle ##

<table> 
 <thead> 
  <tr> 
   <th>监控类型</th> 
   <th>监控项</th> 
   <th>说明</th> 
   <th>级别</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>perf dump throttle-*</td> 
   <td>val</td> 
   <td>当前可用的值</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>max</td> 
   <td>最大限制数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>get</td> 
   <td>获取到的值</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>get_sum</td> 
   <td>获取到的总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>get_or_fail_fail</td> 
   <td>获取或者错误值</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>get_or_fail_success</td> 
   <td>获取或者错误成功值</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>take</td> 
   <td>接受值</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>take_sum</td> 
   <td>接受总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>put</td> 
   <td>推送值</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>put_sum</td> 
   <td>推送总数</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>wait.avgcount</td> 
   <td>等待平均数量</td> 
   <td>&nbsp;</td> 
  </tr> 
  <tr> 
   <td>*</td> 
   <td>wait.sum</td> 
   <td>等待总数</td> 
  </tr> 
 </tbody> 
</table>