One Linux server load up to 600 more exception handling process

# One server load value the process of the hurricane and thought processing time: 2016_11_22:17:00

#收到报警78服务器load值报警, log in to machine uptime view load value

650) this.width=650; "Src=" Http://s5.51cto.com/wyfs02/M00/8A/92/wKiom1g0DkXTtzyJAABBw6D_lOI162.png-wh_500x0-wm_3 -wmp_4-s_2298889790.png "title=" Ce2b8634-faf7-454b-8b16-fccdca5c04dc.png "style=" Float:none; "alt=" Wkiom1g0dkxttzyjaabbw6d_loi162.png-wh_50 "/>

#load值超高, the first reaction top, want to find out which program consumes a lot of CPU

650) this.width=650; "Src=" Http://s2.51cto.com/wyfs02/M00/8A/8E/wKioL1g0DkWAIR9OAAR7eZozhGY187.png-wh_500x0-wm_3 -wmp_4-s_1517068415.png "style=" Float:none; "title=" 94f056db-b5d9-41a5-a4cc-e2f5172fc045.png "alt=" Wkiol1g0dkwair9oaar7ezozhgy187.png-wh_50 "/>

#没有发现, then Vmstat view Normal, iostat,free-m, everything is OK, there are several operations not

650) this.width=650; "Src=" Http://s5.51cto.com/wyfs02/M01/8A/92/wKiom1g0DkWTQZGpAADFotDhczk686.png-wh_500x0-wm_3 -wmp_4-s_3803479640.png "style=" Float:none; "title=" Dd44c43e-f99b-40ed-b05a-45725c7857ab.png "alt=" Wkiom1g0dkwtqzgpaadfotdhczk686.png-wh_50 "/>

#当这一切都正常的时候人有点懵逼, even going to see how the load value of the source uptime is calculated.

However found the code block and do not understand, continue to view the problem, performed netstat check see there is syn_recv this proves that there is service establishment

The connection failed, so I found this program, and then Netstat-tnlp|grep looked at it. 650) this.width=650; "Src=" Http://s2.51cto.com/wyfs02/M02/8A/92/wKiom1g0D_SjO9GYAAEmK9uWXnc341.png-wh_500x0-wm_3 -wmp_4-s_2094274877.png "title=" Fcda40bc-581b-48ac-9544-8a85b3f94b42.png "style=" Float:none; "alt=" wKiom1g0D_ Sjo9gyaaemk9uwxnc341.png-wh_50 "/>

#开始怀疑这个java程序有问题, then view the log, the day is November 22, and the program log stopped 5:32 on November 20, so determine the program problem.

650) this.width=650; "Src=" Http://s2.51cto.com/wyfs02/M00/8A/92/wKiom1g0DkOCr0AcAADVbrW7rM8259.png-wh_500x0-wm_3 -wmp_4-s_2592312870.png "title=" B6774d92-a78f-4850-842e-4ee4306814ed.png "style=" Float:none; "alt=" Wkiom1g0dkocr0acaadvbrw7rm8259.png-wh_50 "/>

#执行重启tomcat命令, this doesn't have to be said, go into the bin directory shutdown.sh the discovery process is gone startup.sh

And then waited a few minutes for the load to fall down.

650) this.width=650; "Src=" Http://s1.51cto.com/wyfs02/M02/8A/8E/wKioL1g0Dt3he_daAABQlqXR4TA513.png-wh_500x0-wm_3 -wmp_4-s_2934274018.png "title=" Eb4d7f02-8c43-41de-b3e1-81449a4124c0.png "alt=" wkiol1g0dt3he_ Daaabqlqxr4ta513.png-wh_50 "/>

Record the idea and experience of dealing with the fault, said more relaxed, but actually spent a certain amount of time, in the implementation of Netstat before the idea, card for a period of time, are ready to flip through the source code, more awkward, but also did not find a deeper reason.

This article is from the "Xiaofanqie" blog, make sure to keep this source http://xiaofanqie.blog.51cto.com/9400142/1875514

One Linux server load up to 600 more exception handling process