OpenTSDB CLI Tools介绍
OpenTSDB提供了客户端工具,可以直接调用相关命令对OpenTSDB进行操作。客户端工具同开源社区版本保持一致,请参见https://opentsdb.net/docs/build/html/user_guide/cli/index.html。
客户端工具使用方法:
- 登录任意一个Master节点。
- 初始化环境变量。
source /opt/client/bigdata_env
- 如果当前集群已启用Kerberos认证,执行以下命令认证当前用户。如果当前集群未启用Kerberos认证,则无需执行此命令。
kinit MRS集群用户
例如:kinit opentsdbuser
- 执行tsdb命令。例如执行tsdb可以打印出当前opentsdb所支持的所有命令,如,fsck, import, mkmetric, query, tsd, scan, search, uid, version。
tsdb: error: unknown command '' usage: tsdb <command> [args] Valid commands: fsck, import, mkmetric, query, tsd, scan, search, uid, version
创建OpenTSDB指标
创建存入到OpenTSDB中的指标名称,可执行tsdb mkmetric sys.cpu.user命令创建sys.cpu.user。
Start run net.opentsdb.tools.UidManager, args: assign metrics sys.cpu.user metrics sys.cpu.user: [0, 0, 6]
查询OpenTSDB指标
tsdb命令可以获取到当前opentsdb存入的指标,可执行tsdb uid metrics sys.cpu.user命令。
Start run net.opentsdb.tools.UidManager, args: metrics sys.cpu.user metrics sys.cpu.user: [0, 0, 6]
若想获取更多信息,可通过执行命令tsdb uid。
Start run net.opentsdb.tools.UidManager, args: Not enough arguments Usage: uid <subcommand> args Sub commands: grep [kind] <RE>: Finds matching IDs. assign <kind> <name> [names]: Assign an ID for the given name(s). rename <kind> <name> <newname>: Renames this UID. delete <kind> <name>: Deletes this UID. fsck: [fix] [delete_unknown] Checks the consistency of UIDs. fix - Fix errors. By default errors are logged. delete_unknown - Remove columns with unknown qualifiers. The "fix" flag must be supplied as well. [kind] <name>: Lookup the ID of this name. [kind] <ID>: Lookup the name of this ID. metasync: Generates missing TSUID and UID meta entries, updates created timestamps metapurge: Removes meta data entries from the UID table treesync: Process all timeseries meta objects through tree rules treepurge <id> [definition]: Purge a tree and/or the branches from storage. Provide an integer Tree ID and optionally add "true" to delete the tree definition Example values for [kind]: metrics, tagk (tag name), tagv (tag value). --config=PATH Path to a configuration file (default: Searches for file see docs). --idwidth=N Number of bytes on which the UniqueId is encoded. --ignore-case Ignore case distinctions when matching a regexp. --table=TABLE Name of the HBase table where to store the time series (default: tsdb). --uidtable=TABLE Name of the HBase table to use for Unique IDs (default: tsdb-uid). --verbose Print more logging messages and not just errors. --zkbasedir=PATH Path under which is the znode for the -ROOT- region (default: /hbase). --zkquorum=SPEC Specification of the ZooKeeper quorum to use (default: localhost). -i Short for --ignore-case. -v Short for --verbose.
向OpenTSDB指标中导入数据
tsdb命令可以使用“tsdb import”命令批量导入指标数据,可执行如下命令:
- 准备指标数据,如包含如下内容的importData.txt文件。
sys.cpu.user 1356998400 41 host=web01 cpu=0 sys.cpu.user 1356998401 42 host=web01 cpu=0 sys.cpu.user 1356998402 44 host=web01 cpu=0 sys.cpu.user 1356998403 47 host=web01 cpu=0 sys.cpu.user 1356998404 42 host=web01 cpu=0 sys.cpu.user 1356998405 42 host=web01 cpu=0
- 导入指标数据,可执行如下命令tsdb import importData.txt。
Start run net.opentsdb.tools.TextImporter, args: importData.txt 2019-06-26 15:45:22,091 INFO [main] TextImporter: reading from file:importData.txt 2019-06-26 15:45:22,102 INFO [main] TextImporter: Processed importData.txt in 11 ms, 6 data points (545.5 points/s) 2019-06-26 15:45:22,102 INFO [main] TextImporter: Total: imported 6 data points in 0.012s (504.0 points/s)
扫描OpenTSDB的指标数据
tsdb命令可以使用“tsdb query”命令批量查询导入的指标数据,例如执行tsdb query 0 1h-ago sum sys.cpu.user host=web01命令。
Start run net.opentsdb.tools.CliQuery, args: 0 1h-ago sum sys.cpu.user host=web01 sys.cpu.user 1356998400000 41 {host=web01, cpu=0} sys.cpu.user 1356998401000 42 {host=web01, cpu=0} sys.cpu.user 1356998402000 44 {host=web01, cpu=0} sys.cpu.user 1356998403000 47 {host=web01, cpu=0} sys.cpu.user 1356998404000 42 {host=web01, cpu=0} sys.cpu.user 1356998405000 42 {host=web01, cpu=0}
删除录入的OpenTSDB指标
tsdb命令可以使用“tsdb uid delete”命令删除录入的指标及值,例如执行tsdb uid delete metrics sys.cpu.user命令。
Start run net.opentsdb.tools.UidManager, args: delete metrics sys.cpu.user