tcollector 不收集数据。 TSDB 为空答案

【问题标题】：tcollector not collecting data. TSDB is emptytcollector 不收集数据。 TSDB 为空
【发布时间】：2014-04-15 09:56:11
【问题描述】：

我已成功安装并启动并运行在

http://54.72.4.157:4242/

我在我们的一台服务器上运行 tcollector，我确实在 startstop.sh 中设置了主机

TSD_HOST=54.72.4.157

我做了

./startstop start

运行所有统计信息收集器。甚至注意到在 tsdb 控制台日志中

[id: 0x5fc4bb31, /54.184.79.13:60203 => /172.31.14.125:4242] CONNECTED: /54.184.79.13:60203

在我的 tcollector 节点上，

ps axl | grep tcollector

我可以看到

0     0 16796 16795  20   0 183712  8000 poll_s Sl   ?          2:17 /usr/bin/python /home/mithralaya/tcollector/tcollector.py -c /home/mithralaya/tcollector/collectors -H 54.72.4.157 -t host=ip-172-31-12-203 -P /var/run/tcollector.pid
4 65534 16806 16796  20   0  39864  3748 poll_s Ss   ?          0:08 /usr/bin/python /home/mithralaya/tcollector/collectors/0/procstats.py
4 65534 16808 16796  39  19  39700  3380 poll_s SNs  ?          0:07 /usr/bin/python /home/mithralaya/tcollector/collectors/0/procnettcp.py
4 65534 16816 16796  20   0  39648  3240 poll_s Ss   ?          0:00 /usr/bin/python /home/mithralaya/tcollector/collectors/0/iostat.py
4 65534 16818 16796  20   0  39648  3400 poll_s Ss   ?          0:01 /usr/bin/python /home/mithralaya/tcollector/collectors/0/ifstat.py
4 65534 16822 16796  20   0  41848  3676 poll_s Ss   ?          0:05 /usr/bin/python /home/mithralaya/tcollector/collectors/0/netstat.py
4 65534 16824 16796  20   0  39648  3524 poll_s Ss   ?          0:00 /usr/bin/python /home/mithralaya/tcollector/collectors/0/dfstat.py
0     0 26617 26171  20   0   8108   940 pipe_w S+   pts/0      0:00 grep --color=auto tcollector

我在 /var/log/tcollector 上的 tcollector 日志中看不到任何重大错误。最新日志

2014-04-15 08:59:40,630 tcollector[16796] WARNING: haproxy.py: Error: HAProxy is not running
2014-04-15 08:59:55,090 tcollector[16796] INFO: removing redis-stats.py from the list of collectors (by request)
2014-04-15 08:59:55,091 tcollector[16796] INFO: removing nfsstat.py from the list of collectors (by request)
2014-04-15 08:59:55,091 tcollector[16796] WARNING: collector hbase_master.py terminated after 16 seconds with status code 1, marking dead
2014-04-15 08:59:55,091 tcollector[16796] INFO: removing udp_bridge.py from the list of collectors (by request)
2014-04-15 08:59:55,091 tcollector[16796] INFO: removing elasticsearch.py from the list of collectors (by request)
2014-04-15 08:59:55,092 tcollector[16796] INFO: removing zfsiostats.py from the list of collectors (by request)
2014-04-15 08:59:55,092 tcollector[16796] INFO: removing varnishstat.py from the list of collectors (by request)
2014-04-15 08:59:55,092 tcollector[16796] INFO: removing mongo.py from the list of collectors (by request)
2014-04-15 08:59:55,093 tcollector[16796] INFO: removing couchbase.py from the list of collectors (by request)
2014-04-15 08:59:55,093 tcollector[16796] INFO: removing graphite_bridge.py from the list of collectors (by request)
2014-04-15 08:59:55,093 tcollector[16796] INFO: removing zfskernstats.py from the list of collectors (by request)
2014-04-15 08:59:55,094 tcollector[16796] INFO: removing smart-stats.py from the list of collectors (by request)
2014-04-15 08:59:55,094 tcollector[16796] WARNING: collector mysql.py terminated after 16 seconds with status code 1, marking dead
2014-04-15 08:59:55,094 tcollector[16796] WARNING: collector hbase_regionserver.py terminated after 16 seconds with status code 1, marking dead
2014-04-15 08:59:55,095 tcollector[16796] INFO: removing postgresql.py from the list of collectors (by request)
2014-04-15 08:59:55,095 tcollector[16796] INFO: removing haproxy.py from the list of collectors (by request)
2014-04-15 08:59:55,095 tcollector[16796] INFO: removing riak.py from the list of collectors (by request)
2014-04-15 08:59:55,095 tcollector[16796] INFO: removing zookeeper.py from the list of collectors (by request)
2014-04-15 08:59:55,096 tcollector[16796] INFO: removing opentsdb.sh from the list of collectors (by request)
2014-04-15 09:09:40,651 tcollector[16796] INFO: Heartbeat (6 collectors running)
2014-04-15 09:19:41,217 tcollector[16796] INFO: Heartbeat (6 collectors running)
2014-04-15 09:29:41,794 tcollector[16796] INFO: Heartbeat (6 collectors running)
2014-04-15 09:39:43,586 tcollector[16796] INFO: Heartbeat (6 collectors running)

但是没有收集到任何统计数据。在 hbase 中，tsdb 和 tsdb-uid 都是空的。

hbase(main):002:0> scan 'tsdb'
ROW                                                          COLUMN+CELL                                                                                                                                                                      
0 row(s) in 0.2890 seconds

hbase(main):003:0>

你也可以在这里看到，

http://54.72.4.157:60010/

如果有人可以帮助我，将不胜感激。

所有基于 Hadoop 的技术都很难安装和配置。我花了一周的时间来解决这个问题，我正在运行 tcollector 24 小时，但 TSDB 中仍然没有数据。

非常感谢，

卡提克

【问题讨论】：

标签： java logging hbase monitoring opentsdb

【解决方案1】：

从日志文件输出看来，实际上没有任何 tcollector 插件在运行。由于错误，它们会生成并在之后立即删除。

【讨论】：

【解决方案2】：

嗯，可能有几个收集器像 procstats.py（它收集 CPU、内存等基本指标）一样运行，我注意到错误日志中没有。

您没有将数据传输到您的 hbase 可能是因为您的 opentsdb 配置设置为您需要手动创建指标的默认值。如果是这样，那么您必须自己定义指标。

相反，要创建自动创建的指标，请尝试转到您的 opentsdb 服务器并检查配置并将指标创建设置为自动。

具体来说，在/etc/opentsdb/opentsdb.conf中，将参数“tsd.core.auto_create_metrics”设置为true，然后重启opentsdb服务。

然后再次检查你的 hbase，看看你是否可以看到 'tsdb-uid' 中的数据。

【讨论】：

【解决方案3】：

请尝试将conf文件设置为自动创建指标

# --------- CORE ----------
# Whether or not to automatically create UIDs for new metric types, default
# is False
tsd.core.auto_create_metrics = true

【讨论】：

我们要添加哪个文件名这一行