cadvisor+influxdb 增加estab统计

内容导读

互联网集市收集整理的这篇技术教程文章主要介绍了cadvisor+influxdb 增加estab统计，小编现在分享给大家，供广大互联网技能从业者学习和参考。文章包含7533字，纯文字阅读大概需要11分钟。

内容图文

1、需求说明：

采用cadvisor+influxdb+grafana进行容器监控并数据展示，同时收集tcpstats相关数据。

启动时关闭默认的—disable-metric，启动参数如下：

[program:cadvisor]
command=/root/go/src/cadvisor/cadvisor -port=18080 -logtostderr=true -v=5 -enable_load_reader=true -storage_duration=15s -disable_metrics="" -docker_only=true -storage_driver=influxdb -storage_driver_db=influxdb -storage_driver_user=influxdb -storage_driver_password=influxdb -storage_driver_host="127.0.0.1:18086"
numprocs=1
autostart=true
autorestart=true
startsecs=3
startretries=5
stopasgroup=true
killasgroup=true
stdout_logfile=/var/log/supervisor/cadvisor_out.log
stderr_logfile=/var/log/supervisor/cadvisor_err.log

2、问题定位

523823198@qq.com

启动后发现向influxdb中写入的数据中并没有tcpstat相关值，so 查看源码如下：

cat storage/influxdb/influxdb.go
...
 50 // Series names
 51 const (
 52     // Cumulative CPU usage
 53     serCpuUsageTotal  string = "cpu_usage_total"
 54     serCpuUsageSystem string = "cpu_usage_system"
 55     serCpuUsageUser   string = "cpu_usage_user"
 56     serCpuUsagePerCpu string = "cpu_usage_per_cpu"
 57     // Smoothed average of number of runnable threads x 1000.
 58     serLoadAverage string = "load_average"
 59     // Memory Usage
 60     serMemoryUsage string = "memory_usage"
 61     // Working set size
 62     serMemoryWorkingSet string = "memory_working_set"
 63     // Cumulative count of bytes received.
 64     serRxBytes string = "rx_bytes"
 65     // Cumulative count of receive errors encountered.
 66     serRxErrors string = "rx_errors"
 67     // Cumulative count of bytes transmitted.
 68     serTxBytes string = "tx_bytes"
 69     // Cumulative count of transmit errors encountered.
 70     serTxErrors string = "tx_errors"
 71     // Filesystem device.
 72     serFsDevice string = "fs_device"
 73     // Filesystem limit.
 74     serFsLimit string = "fs_limit"
 75     // Filesystem usage.
 76     serFsUsage string = "fs_usage"
 77 )
...
...
180 func (self *influxdbStorage) containerStatsToPoints(
181     cInfo *info.ContainerInfo,
182     stats *info.ContainerStats,
183 ) (points []*influxdb.Point) {
184     // CPU usage: Total usage in nanoseconds
185     points = append(points, makePoint(serCpuUsageTotal, stats.Cpu.Usage.Total))
186
187     // CPU usage: Time spend in system space (in nanoseconds)
188     points = append(points, makePoint(serCpuUsageSystem, stats.Cpu.Usage.System))
189
190     // CPU usage: Time spent in user space (in nanoseconds)
191     points = append(points, makePoint(serCpuUsageUser, stats.Cpu.Usage.User))
192
193     // CPU usage per CPU
194     for i := 0; i < len(stats.Cpu.Usage.PerCpu); i++ {
195         point := makePoint(serCpuUsagePerCpu, stats.Cpu.Usage.PerCpu[i])
196         tags := map[string]string{"instance": fmt.Sprintf("%v", i)}
197         addTagsToPoint(point, tags)
198
199         points = append(points, point)
200     }
201
202     // Load Average
203     points = append(points, makePoint(serLoadAverage, stats.Cpu.LoadAverage))
204
205     // Memory Usage
206     points = append(points, makePoint(serMemoryUsage, stats.Memory.Usage))
207
208     // Working Set Size
209     points = append(points, makePoint(serMemoryWorkingSet, stats.Memory.WorkingSet))
210
211     // Network Stats
212     points = append(points, makePoint(serRxBytes, stats.Network.RxBytes))
213     points = append(points, makePoint(serRxErrors, stats.Network.RxErrors))
214     points = append(points, makePoint(serTxBytes, stats.Network.TxBytes))
215     points = append(points, makePoint(serTxErrors, stats.Network.TxErrors))
216     self.tagPoints(cInfo, stats, points)
217
218     return points
219 }
...

源码中并未收集tcpstats的metric进入influxdb，所以需要手动添加。

3、问题解决

修改原代码，将需要的数据添加到points中。

vim storage/influxdb/influxdb.go 
...
 50 // Series names
 51 const (
 52     // Cumulative CPU usage
 53     serCpuUsageTotal  string = "cpu_usage_total"
 54     serCpuUsageSystem string = "cpu_usage_system"
 55     serCpuUsageUser   string = "cpu_usage_user"
 56     serCpuUsagePerCpu string = "cpu_usage_per_cpu"
 57     // Smoothed average of number of runnable threads x 1000.
 58     serLoadAverage string = "load_average"
 59     // Memory Usage
 60     serMemoryUsage string = "memory_usage"
 61     // Working set size
 62     serMemoryWorkingSet string = "memory_working_set"
 63     // Cumulative count of bytes received.
 64     serRxBytes string = "rx_bytes"
 65     // Cumulative count of receive errors encountered.
 66     serRxErrors string = "rx_errors"
 67     // Cumulative count of bytes transmitted.
 68     serTxBytes string = "tx_bytes"
 69     // Cumulative count of transmit errors encountered.
 70     serTxErrors string = "tx_errors"
 71     // Filesystem device.
 72     serFsDevice string = "fs_device"
 73     // Filesystem limit.
 74     serFsLimit string = "fs_limit"
 75     // Filesystem usage.
 76     serFsUsage string = "fs_usage"
 77     // Tcp Establisd count.
 78     serEsTabs string = "tcp_estab"
 79     // Tcp TimeWait count.
 80     serTimeWait string = "tcp_timewait"
 81     // Tcp CloseWait count.
 82     serCloseWait string = "tcp_closewait"
 83 )
...
...
180 func (self *influxdbStorage) containerStatsToPoints(
181     cInfo *info.ContainerInfo,
182     stats *info.ContainerStats,
183 ) (points []*influxdb.Point) {
184     // CPU usage: Total usage in nanoseconds
185     points = append(points, makePoint(serCpuUsageTotal, stats.Cpu.Usage.Total))
186
187     // CPU usage: Time spend in system space (in nanoseconds)
188     points = append(points, makePoint(serCpuUsageSystem, stats.Cpu.Usage.System))
189
190     // CPU usage: Time spent in user space (in nanoseconds)
191     points = append(points, makePoint(serCpuUsageUser, stats.Cpu.Usage.User))
192
193     // CPU usage per CPU
194     for i := 0; i < len(stats.Cpu.Usage.PerCpu); i++ {
195         point := makePoint(serCpuUsagePerCpu, stats.Cpu.Usage.PerCpu[i])
196         tags := map[string]string{"instance": fmt.Sprintf("%v", i)}
197         addTagsToPoint(point, tags)
198
199         points = append(points, point)
200     }
201
202     // Load Average
203     points = append(points, makePoint(serLoadAverage, stats.Cpu.LoadAverage))
204
205     // Memory Usage
206     points = append(points, makePoint(serMemoryUsage, stats.Memory.Usage))
207
208     // Working Set Size
209     points = append(points, makePoint(serMemoryWorkingSet, stats.Memory.WorkingSet))
210
211     // Network Stats
212     points = append(points, makePoint(serRxBytes, stats.Network.RxBytes))
213     points = append(points, makePoint(serRxErrors, stats.Network.RxErrors))
214     points = append(points, makePoint(serTxBytes, stats.Network.TxBytes))
215     points = append(points, makePoint(serTxErrors, stats.Network.TxErrors))
216     points = append(points, makePoint(serEsTabs, stats.Network.Tcp.Established))
217     points = append(points, makePoint(serTimeWait, stats.Network.Tcp.TimeWait))
218     points = append(points, makePoint(serCloseWait, stats.Network.Tcp.CloseWait))
219
220     self.tagPoints(cInfo, stats, points)
221
222     return points
223 }
...

4、数据验证：

相关代码已经修改完成并提交到我的个人仓库中，需要的可以直接拉取进行make build，也可以直接下载cadvisor_with_tcp这个二进制文件执行。

代码仓库：https://github.com/mmjl/cadvisor.git

至此，cadvisor+influxdb，无法收集tcpEstab数量问题已经解决。

cadvisor+influxdb 增加estab统计

标签：tps   ice   mit   make   rest   ecs   容器   ++   append   

本文系统来源：https://blog.51cto.com/momolinux/2428827

内容总结

以上是互联网集市为您收集整理的cadvisor+influxdb 增加estab统计全部内容，希望文章能够帮你解决cadvisor+influxdb 增加estab统计所遇到的程序开发问题。 如果觉得互联网集市技术教程内容还不错，欢迎将互联网集市网站推荐给程序员好友。

内容备注

版权声明：本文内容由互联网用户自发贡献，该文观点与技术仅代表作者本人。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌侵权/违法违规的内容， 请发送邮件至 gblab@vip.qq.com 举报，一经查实，本站将立刻删除。

内容手机端

---