如何获取“桶 ID”如果我有小型转储文件或异常结构，winqual 会使用什么？ (Windows c++)
SO 有一些相关的问题 - 但还没有找到答案 -
我想生成一个“签名”/桶 ID 来向我们的问题跟踪系统报告小型转储/崩溃。由于 MS 已经使用“bucket ids”做到了这一点，我想我可以重复使用他们的bucket/签名生成。
我可以从顶级过滤器或过滤器内的 _EXCEPTION_POINTERS 对象或 _MINIDUMP_EXCEPTION_INFORMATION 结构或从小型转储本身获取该 ID 吗？
这是一个 C++ 应用程序。
There are a few related questions on SO - but have not found the answer -
I would like to generate a "signature"/bucket id to report a minidump/crash back to our issue tracking system. Since MS already does this with "bucket ids" I figured I could just re-use their bucket/signature generation.
Can I get that ID from either a top level filter or the _EXCEPTION_POINTERS object I have inside the filter or the _MINIDUMP_EXCEPTION_INFORMATION structure or from the minidump itself?
This is a C++ application.
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（2）
有一篇论文讨论了 Microsoft 人员如何首先创建存储桶 id（是的，当然，这篇论文来自 Microsoft）。您实际上不需要对他们使用的完全相同的分桶算法进行逆向工程。这个想法行之有效。
该论文可以在 http://www.sigops.org/sosp 找到/sosp09/papers/glerum-sosp09.pdf
他们还制作了一张幻灯片，因为这是一篇 SIGOPS 论文： http:// /www.sigops.org/sosp/sosp09/slides/glerum-slides-sosp09.pdf
There's a paper talking about how Microsoft guys create bucket id at the first place (, and yes, of course, this paper is from Microsoft). You don't really need to reverse engineering exactly the same bucketing algorithm they use. It's the idea that works.
The paper can be found at http://www.sigops.org/sosp/sosp09/papers/glerum-sosp09.pdf
They also do a slide since this is a SIGOPS paper: http://www.sigops.org/sosp/sosp09/slides/glerum-slides-sosp09.pdf
不确定您需要哪个存储桶 ID。数字id由服务器分配，您可以在报告发送后通过查看事件日志找到它。在应用程序事件日志中查找事件 ID 1001。
要查找字符串格式的存储桶 ID，请使用 Windbg 或 cdb 并使用 !analyze 命令。
请注意，出于实际存储桶的目的，您可能需要查看 WATSON_STAGEONE_URL 而不是 BUCKET_ID。
Not sure which bucket id you need. The numeric id is assigned by the server and you can find it after the report is sent by looking at the event log. Look for event ID 1001 in the application event log.
To find the bucket id in string format use windbg or cdb and use the !analyze command.
Note that for the purpose of actually bucketing you will probably want to look at WATSON_STAGEONE_URL rather than BUCKET_ID.