v0.3: bugfix and enhancement

Thanks to all contributors, all those who submitted issues, and reviewers.

• [api] Fix SQL injection (#733) @codingGirlZJ
• Reduce docker image size using multi-stage building (#671) @leikao
• Support for calling hbs with timeout in judge (#251) @qlyzpqz
• [new feature] Support for accessing the redis with a password (#734) @binbing
• Add a new api to update specific user (#710) @danieldai
• [new feature] Support for passing args in plugin scripts (#672) @danieldai
• Compatible with Grafana v5.4.3 (#690) @gcy3y
• [agent] Make metric MemAvailable more accurate in CentOS 7 (#662) @feiguoL
• Ignore the hosts which maintained when aggregating metrics (#609) @YouZhengChuan
• Fix the lastUpdate cache policy in agent when reporting information to hbs (#469) @yangbinnnn
• [new feature] Support for 3-sigma deviation function in judge (#601) @crisewng
• [agent] Fix the panic of agent when running script fail (#468) @yangbinnnn
• [agent] Add new metrics of net.if.xxx.bits (#475) @yangxiangyu
• [agent] Support for monitoring of PCI-E SSD Card (#440) @ElevenShen
• [agent] Support for monitoring of GPU (#425) @sumaig
• [agent] Support for monitoring of flash device (#311) @hongzhenglin
• [api] Support for reverse selecting counters using ! as a prefix of counter
  (#414) @yuline
• [graph] Optimize the IO performance by using multi IO workers (#381) @GaoJiasheng @n4mine
• Optimize the functions such as Md5, PK, PK2, UUID, SortedTags, DictedTagstring, etc. (#393) @n4mine
• Fix crash when giving empty func in judge (#347) @missuzhang
• Optimize memory usage of judge (#326) @yaoxing
• [agent] Fixed an issue where errors were not returned after rpc timed out. (#291) @YouZhengChuan
• [graph] Optimize performance by using + instead of Sprintf when concat string (#287) @n4mine
• [api] Add a new api to partial-update hosts in a host-group (#227) @jojohappy
• [api] Fix a bug that the right value of a strategy can not be floating number (#233) @milkwine

v0.2.1

• #186 [bugfix][api] 修复创建screen或者clone screen 出现「record not found」的问题；
• #149 [bugfix][api] 修复hostgroup_controller中
  DeleteHostGroup的权限问题；
• #150 [bugfix][api] 修复当grafana graph有多个查询时，因为其中一个未匹配到导致其他正常的查询也无法渲染视图；
• #151 [bugfix][api] 修复api配置文件中数据库配置项错别字；
• #173 [bugfix][api] 修复strategy右值判定出错的问题；
• #171 [bugfix][agent] 修复falcon-agent --check panic 的问题；
• #187 [bugfix][graph] 修复graph兼容task和anteye模块；
• #185 [enhancement][hbs] 增加数据库最大连接数配制，减少数据库访问;
• #184 [enhancement][hbs] 策略执行时间支持跨24小时；
• #212 增加透明版本的logo；
• #217 fix(agent): ignore mount point when inodes=0 ;
• #219 [enhancement][agent] support get agent endpoint from ENV ENDPOINT;
• #132 #144 [enhancement][api] 完善api；

open-falcon v0.2.0 release

统一的后端

• alarm支持报警历史信息入库存储和展示；
• 「报警合并」模块links的功能合并到统一前端 Dashboard 中，降低用户配置和维护成本；
• 「报警发送」模块sender的功能合并到 alarm 中，降低用户配置和维护成本；
• query的功能合并到了falcon-api组件中；
• 支持非周期性上报数据存储；
• agent支持通过自定义配置，只采集指定磁盘挂载点的磁盘监控数据；
• agent支持配置一个默认 tag，这样通过该 agent 上报的所有数据都会自动追加这个tag；
• judge新增报警判断函数lookup(#num, limit)，如果检测到过去num个周期内，有limit次符合条件就报警；

bugfix

• 修复grafana不支持metric含有大写字母的bug；
• 修复agent写多个transfer高可用不生效的bug；
• 修复agent发送数据给transfer的超时设置不合理的问题；

全新的 RESTful API：让 open-falcon 没有难自动化的操作

• 发布了全新设计的组件 falcon-api，falcon-plus 所有的功能都可以通过 RESTful API 来完成；
• 统一前端 Dashboard 绝大部分功能都是通过 falcon-plus api 来实现；