feat: implement snowflake algorithm

[azhsmesos]

解决两个问题：
1.时钟回拨，没有采用redis的方式，而是本地保存当前最大timestart，每次启动时就找到这台机器上面的max_time自增就好了，保持逻辑时钟
2.跨毫秒起始值起始每次默认都是0，这会导致末尾是偶数不方便取余，所以采用random的方式选取1～3的随机数

[codecov]

Codecov Report

Merging #659 (1c2602b) into main (00a5984) will decrease coverage by 0.13%.
The diff coverage is 40.90%.

@@            Coverage Diff             @@
##             main     #659      +/-   ##
==========================================
- Coverage   60.92%   60.78%   -0.14%     
==========================================
  Files         120      121       +1     
  Lines        6377     6421      +44     
==========================================
+ Hits         3885     3903      +18     
- Misses       2120     2142      +22     
- Partials      372      376       +4     

Impacted Files	Coverage Δ
...ponents/sequencer/snowflake/sequencer_snowflake.go	40.90% <40.90%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 2aa9090...1c2602b. Read the comment docs.

[seeflood]

need to implement GetSegment

[seeflood]

action:
调研一下要不要接存储
调研一下时钟回拨问题
沟通一下GLCC 课题，要先提交申请再写PR

@ZLBer @seeflood

[seeflood]

Need english comments

[ZLBer]

@seeflood
This subject participated in GLCC.I think we should follow the rules of GLCC.
PR should not be submitted now.It's not fair.
We should choose the candidate from all students who have submitted their proposal.

[seeflood]

@ZLBer Agreed. We should follow the rules and make it fair.
@azhsmesos Thanks for your contribution, but please submit a GLCC proposal and wait for the approval before submitting this PR.
We have to close this PR and wait for the GLCC process.
For now, you can continue with other PR, like #554 and the istio one
Thanks for your understanding, and sorry for the inconvenience !