Skip to content

Commit 5e0fc9c

Browse files
committed
update standups
1 parent 389eb01 commit 5e0fc9c

File tree

2 files changed

+72
-19
lines changed

2 files changed

+72
-19
lines changed

decisions/0040-agile-daily-standup-meeting.md

+18-19
Original file line numberDiff line numberDiff line change
@@ -4,37 +4,36 @@ Date: 2017-06-16
44

55
## Status
66

7-
Accepted
7+
Proposed
88

99
## Context
1010

11-
effective
12-
purpose, comunication, sharing planning today, what done yesterday.
13-
1. Purpose
14-
2. Howto
15-
3. Stand up
16-
4. Board(stand around)
17-
5. One minute per person
18-
6.
11+
1. Team `devops` set up recently, we need some change to improve the performance;
12+
2. No communication in the team, always person to person;
13+
3. Tasks not organized well which spread among different projects;
14+
4. Some teams require daily reports for their management, but lose the communication between team members.
1915

2016
## Decision
2117

22-
Effective standups are all about communication and commitment, not about status updates.
18+
We're going to have 10 minutes daily standup meetings with each team so that We're on the same page about what everyone in the team is working on and we can strategize on how to approach the work items on the board for the day to come.
2319

24-
* Less than 15 minutes;
25-
* Everyone should standup;
26-
* Safe time - make sure everyone free that time;
27-
* Focus - Every should watching the board;
28-
We're on the same page about what everyone in the team is working on and we can strategize on how to approach the work items on the board for the day to come.
29-
30-
# Meeting
31-
32-
So before every scrum take 5 minutes to actually think and note down what you were planning to accomplish yesterday, did you actually accomplish what you've committed to, if not why?
20+
Effective standups are all about communication and commitment, not about status updates. So before every scrum take 5 minutes to actually think and note down what you were planning to accomplish yesterday, did you actually accomplish what you've committed to, if not why?
3321

3422
Take some time to plan your day and decide what you're going to accomplish today, commit to it and state it to the team (not the management), so the team members will keep each other accountable for the commitments they make to make the whole team improve together and make sure to remove impediments that prevent team members to accomplish what they plan.
3523

3624
Scrum master should pay more attention to the status of each card, encourage team members to deploy in time.
3725

26+
### Keynotes
27+
28+
* Safe time - make sure everyone free that time;
29+
* Everyone should standup;
30+
* Focus - Every should watching the board;
31+
* Less than 10 minutes for the whole meeting;
32+
* One minute per person, no more than two minutes;
33+
* Remove impediments;
34+
* What done yesterday;
35+
* What will do today.
36+
3837
## Consequences
3938

4039
Refs:
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,54 @@
1+
# 41. The summary of MySQL to RDS
2+
3+
Date: 2017-07-17
4+
5+
## Status
6+
7+
Accepted
8+
9+
## Context
10+
11+
1. 当前数据库为 Master, Slave 模式;
12+
2. Master 实例上创建了 16 个数据库,全部需要迁移;
13+
3. 部分业务存在跨库查询;
14+
4. 有一个数据库被近 10 个业务进行查询等数据库操作;
15+
5. 当前数据库为 MySQL 5.1,存在主从同步性能问题;
16+
6. RDS 仅支持 MySQL 5.5, 5.6, 5.7 版本;
17+
18+
## Decision
19+
20+
1. 调研 MySQL 5.1 与 5.5 之间的差异,并周知各个项目组成员;
21+
2. 将数据库与业务之间的引用关系用二维表列举出来;
22+
3. 选出被业务单独或少量引用的数据库,先通过将对应测试数据库进行迁移,并交由测试人员进行测试;
23+
4. 测试完成后对正式环境直接进行迁移;
24+
5. 针对被10多个业务引用的数据库迁移,我们不止要做测试环境的迁移,还要做线上环境的测试
25+
1. 为了保证线上测试可回滚,我们需限定只有测试人员进行操作;
26+
1. 限制我们的服务器安全组,外网只能通过办公网络访问;
27+
2. 限制我们的入口 slb,只能通过办公网络访问。
28+
2. 我们的服务之间存在调用关系,部分业务走的外网域名;
29+
1. 准备我们所有外网业务的域名及其内网 ip 映射,通过 ansible 分发并追加其至所有的线上机器 hosts
30+
3. 所有业务准备好分支 feature/db-config-update-to-rds,此分支与 master 分支差异只有数据库配置改动,冻结 master 新代码合入,如此确保随时可上线,随时可回滚;
31+
4. 创建数据库迁移任务: 结构迁移+整体数据+增量数据迁移;
32+
5. 停止并备份任务计划调度(/etc/crontab,cron.d,/var/spool/cron);
33+
6. 停止所有业务服务;
34+
7. 停止所有 nginx;
35+
8. 脚本验证所有机器上的服务和任务计划,确保无运行中相关程序;
36+
9. 验证无数据库连接;
37+
10. 锁定原数据库写操作(验证确认确实不可写)
38+
11. 网络隔离;
39+
12. 检查源数据库与目的数据库 数据一致性;
40+
13. 停止数据迁移任务;
41+
14. 停止主db数据库服务;
42+
15. 启动RDS到从库的数据同步任务;
43+
16. 重启所有服务;
44+
17. 重启 nginx;
45+
18. 测试团队开始全面验证;
46+
19. 网络隔离解除;
47+
20. 恢复任务计划;
48+
21. 重新执行停服期间计划任务。
49+
50+
## Consequences
51+
52+
1. RDS 磁盘预留过低,导致同步中途停止,无法进行写操作;
53+
2. 一些业务需要回调,测试不完整;
54+
3. 如果有完整的预发布环境,可保证服务零停机时间;

0 commit comments

Comments
 (0)