https://yq.aliyun.com/articles/647468 基本用法 pstack pid strace -e trace=all -T -tt -p pid strace -o server.strace -Ttt -p pid 应用举例 python while True: print('hello') 查看应用进程 ps -ef | grep python root 13491 13054 13 13:27 pts/0 00:00:06 python pstack 13491 #0 0x0000726388f9ea90 in __write_nocancel () at ../sysdeps/unix/syscall-template.S:81 #1 0x0000726388f292f3 in _IO_new_file_write (f=0x726389277400 <_IO_2_1_stdout_>, data=0x72638a18d000, n=2) at fileops.c:1258 #2 0x0000726388f2ab0e in new_do_write (to_do=2, data=0x72638a18d000 "1\n. \b[8@while True:\033[C", fp=0x726389277400 <_IO_2_1_stdout_>) at fileops.c:519 #3 _IO_new_do_write (fp=0x726389277400 <_IO_2_1_stdout_>, data=0x72638a18d000 "1\n. \b[8@while True:\033[C", to_do=2) at fileops.c:495 #4 0x0000726388f29a50 in _IO_new_file_xsputn (f=0x726389277400 <_IO_2_1_stdout_>, data=<optimized out>, n=1) at fileops.c:1326 #5 0x0000726388f1e05b in __GI__IO_fputs (str=0x726389cd2a8d "\n", fp=0x726389277400 <_IO_2_1_stdout_>) at iofputs.c:39 #6 0x0000726389c0b9c0 in PyFile_WriteString () from /lib64/libpython2.7.so.1.0 #7 0x0000726389c80d48 in PyEval_EvalFrameEx () from /lib64/libpython2.7.so.1.0 #8 0x0000726389c8908d in PyEval_EvalCodeEx () from /lib64/libpython2.7.so.1.0 #9 0x0000726389c89192 in PyEval_EvalCode () from /lib64/libpython2.7.so.1.0 #10……

安装 yum install gdb 使用 pstack pid gstack pid pstree……

ERROR 2013 (HY000) Lost connection to MySQL server during query 错误出现场景 select count(id) from account_user where id> 0; +----------+ | count(id) | +----------+ | 2940245 | +----------+ 1 row in set (0.78 sec) delete from account_user where id> 0; ERROR 2013 (HY000): Lost connection to MySQL server during query tidb.log 2018/11/14 10:57:12.476 server.go:303: [info] con:261 new connection 10.1.88.32:54462 2018/11/14 10:57:15.387 coprocessor.go:689: [info] [TIME_COP_PROCESS] resp_time:792.561353ms txn_start_ts:404269101106331649 region_id:77367 store_addr:10.1.88.84:20160 kv_process_ms:537 2018/11/14 10:57:17.227 pd.go:107: [warning] get timestamp too slow: 135.012109ms 2018/11/14 10:57:17.812 pd.go:107: [warning] get timestamp too slow: 56.316863ms 2018/11/14 10:57:17.821 coprocessor.go:689: [info] [TIME_COP_PROCESS] resp_time:3.226284139s txn_start_ts:404269101106331649 region_id:77863 store_addr:10.1.88.85:20160 kv_process_ms:1580 scan_total_write:314161 scan_processed_write:314160 scan_total_data:314160 scan_processed_data:314160 scan_total_lock:1 scan_processed_lock:0 2018/11/14 10:57:17.826 coprocessor.go:689: [info] [TIME_COP_PROCESS] resp_time:3.231259247s txn_start_ts:404269101106331649 region_id:77590 store_addr:10.1.88.85:20160 kv_process_ms:1807 scan_total_write:355885 scan_processed_write:355884 scan_total_data:355884 scan_processed_data:355884 scan_total_lock:1 scan_processed_lock:0 2018/11/14 10:57:17.872 coprocessor.go:689: [info] [TIME_COP_PROCESS] resp_time:3.276994472s txn_start_ts:404269101106331649 region_id:77576 store_addr:10.1.88.86:20160 kv_process_ms:1684 scan_total_write:346208 scan_processed_write:346207 scan_total_data:346207 scan_processed_data:346207 scan_total_lock:1 scan_processed_lock:0 2018/11/14……

写在前面 在学习Service Mesh前还是最好先清楚几个概念，如熔断，限流，降级。接下来一起揭开面纱，让陌生的名词变得没有那么神秘。一切都来自于现实的应用场景，及解决问题的方法方式。 consumer表示服务调用方 provider表示服务提供方 A调用B服务，一般是泛指调用B服务里面……

Service Mesh(服务网格) Kubernetes 已经给我们带来了诸多的好处。但是仍有些需求比如 A/B 测试、金丝雀发布、限流、访问控制,端到端认证等需要运维人员进一步去解决。 Istio 是完全开源的服务网格,提供了一套完整的解决方案，可以透明地分层到现有的分布式应用程序上。对开发人员几乎无感的同时获得超能力。 如果想要现……

https://segmentfault.com/a/1190000011174694……

蓝绿发布 准备两套环境，蓝和绿。绿环境为当前正式环境，现在将新版部署到蓝环境当中并进行测试。没有问题后将路由指向蓝环境。若发现问题可将路由指回到原来的绿环境中进行回滚。 问题： 如果是无状态应用还可以，如果应用有状态或类似数据库之类的应该很麻烦，还有新旧版本是否兼容的问题。 A/B 测试 用来测……

安装 ubuntu14.04 .安装kvm apt-get install qemu-kvm libvirt0 virtinst bridge-utils virt-viewer .配置实体机网络 cat /etc/network/interfaces auto lo iface lo inet loopback auto br0 iface br0 inet static address 10.0.*.* netmask 255.255.0.0 gateway 10.1.0.1 type bridge bridge_ports eth0 dns-nameservers 114.114.114.114 .创建一个虚拟机 virt-install --connect qemu:///system -n test01 -r 1024 -f /home/kvm/test01.qcow2 -s 20 -c /home/kvm/ubuntu-12.04.1-server-amd64.iso --vnc --noautoconsole --os-type linux --os-variant ubuntuPrecise --accelerate --network=bridge:br0 centos7 yum install qemu-kvm libvirt virt-install bridge-utils .配置实体机网络 cat ifcfg-enp7s0 DEVICE="enp7s0" ONBOOT=yes BOOTPROTO=static UUID=96a09db3-9b06-4a50-8d0c-9868cf531b84 #HWADDR="08:60:6E:06:C7:1A" TYPE=Ethernet BRIDGE=br0 IPADDR=0.0.0.0 cat ifcfg-br0 DEVICE=br0 ONBOOT=yes BOOTPROTO=static TYPE=Bridge IPADDR=10.1.*.* PREFIX=16 GATEWAY=10.1.0.1 DNS1=223.5.5.5 查看 brctl show .启动 libvirt systemctl start libvirtd systemctl enable libvirtd .创建虚拟机 virt-install --virt-type kvm --name test01 --ram……

https://www.elastic.co/solutions/apm 应用程序性能监控 整体架构 先搞起来 agent 收集信息 apm-server 接受agent信息并发送到ES ES 存储信息 Kibana 信息检索展示 agent python django 安装扩展包 pip install elastic-apm django 配置 # Add the agent to the installed apps INSTALLED_APPS = ( 'elasticapm.contrib.django', #... ) ELASTIC_APM = { # Set required service name. # Allowed characters: # a-z, A-Z, 0-9, -, _, and space 'SERVICE_NAME': 'my-app', #后台进程 'TRANSPORT_CLASS': 'elasticapm.transport.http.AsyncTransport', # Use if APM Server requires a token #'SECRET_TOKEN': '', # 没有数据可以开启debug查看 # 'DEBUG': True, # Set custom APM Server……

当前连接数 SELECT count(*) FROM pg_stat_activity WHERE NOT pid=pg_backend_pid(); count ------- 3 (1 row) 数据库占用空间 select pg_size_pretty(pg_database_size('postgres')); pg_size_pretty ---------------- 14 MB (1 row) or \l+ 数据库表(不包括索引)或单条索引占用空间 select pg_size_pretty(pg_relation_size('t_name')); pg_size_pretty ---------------- 24 kB (1 行记录) or \d+ 表中所有索引占有的空间 select pg_size_pretty(pg_indexes_size('t_name')); pg_size_pretty ---------------- 280 kB (1 行记录) 表中索引占用的空间大小排序 select indexname , pg_size_pretty(pg_relation_size(indexname::regclass)) ,indexdef from pg_indexes where tablename = 't_name' order by pg_relation_size(indexname::regclass) desc limit 10; 表和索引占用总空间 select pg_size_pretty(pg_total_relation_size('t_name')); pg_size_pretty ---------------- 380 kB (1 行记录) 查……