Skip to content

Commit

Permalink
Format the tags
Browse files Browse the repository at this point in the history
  • Loading branch information
zedware committed Apr 15, 2023
1 parent a4613ff commit d2c5795
Show file tree
Hide file tree
Showing 138 changed files with 270 additions and 206 deletions.
18 changes: 8 additions & 10 deletions _includes/tag_cloud.html
Original file line number Diff line number Diff line change
@@ -1,13 +1,11 @@
<h2>Tag Cloud</h2>
{% capture temptags %}
{% for tag in site.tags %}
{{ tag[1].size | plus: 1000 }}#{{ tag[0] }}#{{ tag[1].size }}
{% endfor %}
{% endcapture %}
{% assign sortedtemptags = temptags | split:' ' | sort | reverse %}
{% for temptag in sortedtemptags %}
{% assign tagitems = temptag | split: '#' %}
{% capture tagname %}{{ tagitems[1] }}{% endcapture %}
<a href="{{ site.tag_page }}{{ tagname }}"><code class="highligher-rouge"><nobr>{{ tagname }}</nobr></code></a>
{% assign tags = site.tags | sort %}
{% for tag in tags %}
<span class="site-tag">
<a href="{{ site.tag_page }}{{ tag | first | slugify }}/"
style="font-size: {{ tag | last | size | times: 4 | plus: 80 }}%">
{{ tag[0] | replace:'-', ' ' }} ({{ tag | last | size }})
</a>
</span>
{% endfor %}

2 changes: 1 addition & 1 deletion _posts/2003-03-04-NULL.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: 空值的问题
original: https://blog.csdn.net/zedware/article/details/20488167
tags: sql
tags: SQL
---

空值的问题[1]()
Expand Down
2 changes: 1 addition & 1 deletion _posts/2003-04-04-UPDATABLE-VIEW.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: 可更新视图
original: https://blog.csdn.net/zedware/article/details/20488069
tags: sql view
tags: SQL VIEW
---

视图是从一个或多个基表(或视图)导出的表。通常视图仅仅是一个虚表,即数据库中只存放视图的定义信息等元数据,而不存放视图对应的数据。 视图上的操作和基表类似,但是 DBMS对视图的更新操作(INSERT、DELETE、UPDATE)往往存在一定的限制。 DBMS对视图进行的权限管理和基表也有所不同。
Expand Down
4 changes: 2 additions & 2 deletions _posts/2004-06-01-DATA-TYPES.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: 数据库中的数据类型
original: https://blog.csdn.net/zedware/article/details/21247193
tags: sql
tags: SQL
---

数据库中的数据类型通常有很多种,也有不同的分类方法。例如最常见的数值型、字符型、日期时间型,也有不太常见的布尔型、枚举型、集合型等。要在DBMS中实现某种具体数据类型(例如最简单的INTEGER)的支持,我们可以从以下几个方面来考虑。
Expand Down Expand Up @@ -41,4 +41,4 @@ tags: sql

对于简单的整数类型,格式主要是字节顺序的问题。但是对于比较复杂的NUMEIRC类型,更重要的是考虑存储和操作效率了。



2 changes: 1 addition & 1 deletion _posts/2004-06-01-SQL-TOPN.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: 用SQL找出前N名
original: https://blog.csdn.net/zedware/article/details/50926404
tags: sql
tags: SQL
---

业务系统中经常会有排名的需求,考试和比赛中则更普遍了。Excel 中也有个 Rank 函数供排名之用,数据库中更不例外了。假设需要找出工资最高的前三个员工工资(及其员工号)。不过,“前三名”的具体含义需要准确的定义,不然查出来的可能不是想要的结果。首先,因为表中记录数可能就少于三,查出来的记录可能等于三条也可能少于三条。其次,需要考虑并列名次的处理。
Expand Down
2 changes: 1 addition & 1 deletion _posts/2004-07-01-SOURCEINSIGHT.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: 在SourceInsight中集成版本管理工具
original: https://blog.csdn.net/zedware/article/details/20487555
tags: source-insight
tags: [SourceInsight]
---

    SourceInsight是一个功能强大的源码阅读和编辑工具,它也支持用户自定义的命令。我们可以选取主菜单中的“View”->“Toolbars”中的“Source Control”和“Build”以便显示版本管理和构建工具栏。
Expand Down
2 changes: 1 addition & 1 deletion _posts/2005-02-16-TYPE-DISCRIMINATION.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
---
layout: post
title: SQL 数据类型辨析
tags: sql
tags: SQL
---

* * *
Expand Down
2 changes: 1 addition & 1 deletion _posts/2005-08-05-ISOLATION-LEVEL.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
---
layout: post
title: 事务的隔离级别(Transaction Isolation Level)
tags: sql
tags: SQL
---

事务隔离级别是对事务之间的隔离性作出的一个度量。它通常以下面的几个现象进行考察。注意这些现象都是从读的角度来分析的,即分析写操作对读操作的影响。
Expand Down
2 changes: 1 addition & 1 deletion _posts/2005-08-31-TPMC-WAREHOUSE.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
---
layout: post
title: tpmC per Warehouse in TPC-C
tags: tpc-c tpmc
tags: TPC-C tpmC
---

As defined in the TPC-C specification, their is an upper bound in the tpmC metric as calculated by the number of warehouses. To get a higher tpmC, vendors have to load more number of warehouses.
Expand Down
2 changes: 1 addition & 1 deletion _posts/2006-05-18-NULL-AGAIN.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
---
layout: post
title: 此NULL非彼NULL
tags: sql
tags: SQL NULL
---

* * *
Expand Down
2 changes: 1 addition & 1 deletion _posts/2006-05-18-ROWNO.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
---
layout: post
title: 从给查询结果集增加行号说起
tags: sql shell
tags: SQL Shell
---

* * *
Expand Down
2 changes: 1 addition & 1 deletion _posts/2006-05-18-SYMLINK.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
---
layout: post
title: 符号链接、硬链接、快捷方式和Junctions
tags: file-system
tags: [File System]
---

* * *
Expand Down
2 changes: 1 addition & 1 deletion _posts/2006-07-17-HELLOWORLD.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
---
layout: post
title: 各种语言写就的“hello world”程序
tags: hello-world
tags: [Hello World]
---

* * *
Expand Down
2 changes: 1 addition & 1 deletion _posts/2006-08-10-STACKOVERFLOW.md
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
---
layout: post
title: What's stack overflow?
tags: stack-overflow
tags: [Stack Overflow]
---

* * *
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: 用尽量少的语句和C语言关键字实现一个函数
original: https://blog.csdn.net/zedware/article/details/1685101
tags: c
tags: C
---

```
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: 微软的数据访问技术(From MDAC to Windows DAC)
original: https://blog.csdn.net/zedware/article/details/4917233
tags: mdac
tags: MDAC
---

在过去的很多年中,微软发展了很多种具体的数据访问技术。然而,随着技术的发展和进步,很多技术现在都显得很过时了;有些技术因为被非常广泛的使用,仍然得到了一定程度的维护和支持。10多年前被视为主流的技术,到今天不是被彻底抛弃就是被视为 Legacy 了。这里是微软的:[Data Access Technologies Road Map](http://msdn.microsoft.com/en-us/library/ms810810.aspx)。它的维护历史为:January 2002;Revised December 2008,这也基本说明了最近的这些年中,被称为 MDAC 的这部分技术没有什么实质的发展,而是处于一个维护状态。
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: MySQL将会像BSD、Linux那样弄出不同的分支和发行版?
original: https://blog.csdn.net/zedware/article/details/5015062
tags: mysql
tags: MySQL
---

    从用户的角度看,MySQL被Sun收购,又被Oracle收购可能真不是什么好事。MySQL本来就被Oracle通过并购其依赖的存储引擎Innobase、BerkeleyDB折腾的不行,又在被Sun收购后发生了[Monty创建新分支MariaDB](http://askmonty.org/wiki/index.php/MariaDB)这样的事情,最终还是随着Sun一起被Oracle并购,没有逃脱Larry Elison的魔爪。也许,开源软件的好处就在于选择比较多,可以适合不同人的口味,而且你觉得不合适的,自己就可以去做些改造。不过对于普通的用户甚至开发者而言,选择太多了也不见得是什么好事——会弄得无所适从,不知道身在何处。
Expand Down
2 changes: 1 addition & 1 deletion _posts/2009-12-23-Google Public DNS.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: Google Public DNS
original: https://blog.csdn.net/zedware/article/details/5058973
tags: dns
tags: DNS
---

   已经有了一个叫做OpenDNS的非Open Source DNS服务,现在又出来了一个号称Public的Google Public DNS。从[http://code.google.com/intl/zh-CN/speed/public-dns/](http://code.google.com/intl/zh-CN/speed/public-dns/)可以看到,Google提供的这两个DNS服务的IP地址很好记:8.8.8.8和8.8.4.4。要是大家都将DNS解析大集中到Google,那好,Google更可以容易的知道你整天都访问了哪些站点了。
2 changes: 2 additions & 0 deletions _posts/2009-12-26-ODBC重新焕发生机.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,9 @@
layout: post
title: ODBC重新焕发生机
original: https://blog.csdn.net/zedware/article/details/5083443
tags: ODBC
---

    ODBC是数据库访问接口中的一员大将。我们可以从SQL标准的SQL/CLI中看到它的身影,也可以从JDBC中看到它的巨大影响。随着数据库技术的发展,ODBC在1990年代中期就已经非常成熟稳定了,在随后的这些年中基本停止了发展。随着微软Windows 7的发布,ODBC又推出了3.8版本,在以下几个方面做出了较大的改进(引用自微软的[What's New in ODBC 3.8](http://msdn.microsoft.com/en-us/library/ee388580(VS.85).aspx)):
- Executing connection operations asynchronously. For more information, see[Asynchronous Execution](http://msdn.microsoft.com/en-us/library/ms713563(VS.85).aspx).

Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -2,8 +2,9 @@
layout: post
title: Oracle Makes Commitments to Customers, Developers and Users of MySQL
original: https://blog.csdn.net/zedware/article/details/5102723
tags: mysql
tags: MySQL
---

[Oracle Makes Commitments to Customers, Developers and Users of MySQL](http://www.marketwire.com/press-release/Oracle-Corporation-NASDAQ-ORCL-1090000.html),半个月前,著名的甲骨文受不了欧盟的反垄断折腾了,做出了这么10条承诺。不过我觉得这么写承诺未必有什么实质的意义,也许只是给欧盟个回应,以后并购后该怎么样还是怎么样。虽然Oracle早几年就视MySQL为一个潜在的威胁,使出了釜底抽薪的把戏,购买了MySQL的存储引擎Innobase和BerkeleyDB;但是,经过Sun折腾过一次的MySQL对于Oracle来说还不是并购的主要目的。

还是人家Monash Research反应迅速,December 14, 2009也就发出了博客:[Oracle lifts the cloud hanging over MySQL storage engine vendors](http://www.dbms2.com/2009/12/14/oracle-mysql-storage-engine/)。比我这个短文要及时、详细多了。MySQL的创始人Monty整了个[“救救MySQL”的全球抗议](http://www.helpmysql.org/cn/petition),并到处散发,希望各大经济体能够重视对Oracle并购案的审查,为MySQL的的未来留下生机。
Expand Down
2 changes: 1 addition & 1 deletion _posts/2010-03-11-表达式的求值顺序.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: 表达式的求值顺序
original: https://blog.csdn.net/zedware/article/details/5371715
tags: sql
tags: SQL
---

在C等高级语言中,对表达式求值时一般都有一套较为固定的规则。例如按照从左到右的顺序求值,如果算到某处可以得出结果就忽略后续的表达式(Short Circuit)。举个例子:
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: 自治事务(Autonomous Transaction)
original: https://blog.csdn.net/zedware/article/details/5521392
tags: sql
tags: SQL
---

    “事务”是一个大家耳熟能详的名词了。事务具有四个基本的特性,我们分别将其称为原子性(Atomic)、一致性(Consistency)、隔离性(Isolation)、持久性(Durability)。统称为ACID。与事务有关的还有“分布式事务”、“超长事务”等正式和非正式的名词。这里要说的是“自治事务”(Autonomous Transaction)。
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: Oracle与MySQL中“可串行化”的对比测试
original: https://blog.csdn.net/zedware/article/details/5529389
tags: mysql oracle
tags: MySQL Oracle
---

Thomas Kyte (Oracle 世界中那个著名的 Tom)在“Oracle 9i&10g编程艺术”一书中为了说明事务的隔离级别,提供了一些例子。这里我们来看看“可串行化”级别下的表现。
Expand Down
2 changes: 1 addition & 1 deletion _posts/2010-08-27-Again, Notes on Alignment.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: Again, Notes on Alignment
original: https://blog.csdn.net/zedware/article/details/5844522
tags: c
tags: C
---

Alignment 是编程时比较普遍的问题。例如采用 C/C++ 之类的语言进行编程时经常要考虑不同 CPU 体系结构带来的字节对齐问题,有时甚至要考虑 cache line 的对齐问题,以避免不必要的性能损失。不幸的是,随着硬盘技术的发展,它也出现了 Alignment 的问题。因为扇区大小从以前大家熟知的 512 Byte 增长到了 4096 Byte:[http://bugs.gentoo.org/show_bug.cgi?id=304727](http://bugs.gentoo.org/show_bug.cgi?id=304727)
Expand Down
2 changes: 1 addition & 1 deletion _posts/2010-09-05-heap mgmt and multi-threaded..md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: heap mgmt and multi-threaded.
original: https://blog.csdn.net/zedware/article/details/5865181
tags: c
tags: C
---

Refereces:
Expand Down
1 change: 1 addition & 0 deletions _posts/2011-01-05-Notes on deduplication.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@ layout: post
title: Notes on deduplication
original: https://blog.csdn.net/zedware/article/details/6118892
---

Data de-duplication是存储中比较成熟的技术了。无论个人的桌面机,还是企业的服务器,或是备份用的存储系统中,都存在大量的重复数据。最简单的一种就是同一个文件存有多份,稍微复杂一点的是类似的文件存在多份。如果是备份系统,则数据的多次全备份之间更是存在着大量的荣誉数据,即便是多次增量备份或差分备份之间,也存在着很多的冗余数据。解决这些冗余数据,减少存储空间,提高恢复效率的办法有很多。最简单也最容易想到的是Single Instance Storage,也就是每个同样的文件仅仅保留一份,而不是重复保留。稍复杂的办法可以借鉴版本管理的办法,对于同一个文件的不同版本仅仅保存差异部分。更复杂的办法是识别不同文件/对象之间的重复部分,将其冗余消除,仅记录一份。不同文件或对象之间重复部分的识别比较复杂,又要保证速度,又要保证廉价,更要保证数据完整性。不过业界已经有了很多成熟的算法和产品,例如Data
Domain。
压缩其实就是一种去除数据重复的办法,只不过压缩的粒度偏小,大家不把它当做典型的数据重复消除技术。从数据重复消除的不同层次看,也许可以将其分为几层:
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@ layout: post
title: Optimizing the Linux kernel for Database?
original: https://blog.csdn.net/zedware/article/details/6169265
---

Oracle is now a big player in the Linux community. Vetical integration is a trend in the computer world.
IBM, Oracle, Microsoft, HP, Dell, Cisco, VMware, EMC and other big players in the IT field are doing great
works in integrating softwares and even hardwares. Different layers of system share the same problems
Expand Down
1 change: 1 addition & 0 deletions _posts/2011-02-16-DLL hell again - this time on Linux.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@ layout: post
title: DLL hell again - this time on Linux
original: https://blog.csdn.net/zedware/article/details/6189219
---

突然发现Fedora中的yum不好使了,一执行就报错,说什么不能import某某库。
折腾半天,发现曾经手工装过Python,莫非是它搞出了包的冲突?

Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@ layout: post
title: Collections of Data Deduplication Links
original: https://blog.csdn.net/zedware/article/details/6225162
---

1. 入门

在大多数时候,Wikipedia是个迅速了解某个特定技术发展背景的不错选择。例如:[http://en.wikipedia.org/wiki/Data_deduplication](http://en.wikipedia.org/wiki/Data_deduplication)
Expand Down
1 change: 1 addition & 0 deletions _posts/2011-03-07-Bloom Filter in Oracle Database.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,5 +3,6 @@ layout: post
title: Bloom Filter in Oracle Database
original: https://blog.csdn.net/zedware/article/details/6230135
---

Bloom Filter用的范围很广,不仅在Search Engine中大量采用,在Data Deduplication中也频频遇到。那么它在数据库中的应用情况怎么样呢?放狗找一下,这个比较有意思:[http://antognini.ch/papers/BloomFilters20080620.pdf](http://antognini.ch/papers/BloomFilters20080620.pdf)
今天在和同事讨论问题的时候突然想起来,这个玩意就是以前在Database Machine论文中见过的Bit vector啊:http://academic.research.microsoft.com/Publication/803587
2 changes: 1 addition & 1 deletion _posts/2011-04-09-PORTABLE.md
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
layout: post
title: 可移植性真的很重要:on shell and tr
original: https://blog.csdn.net/zedware/article/details/6311400
tags: shell
tags: Shell
---

道理很简单,但是不小心造成的麻烦会很大!
Expand Down
2 changes: 1 addition & 1 deletion _posts/2011-06-01-GNUDOT.md
Original file line number Diff line number Diff line change
Expand Up @@ -65,4 +65,4 @@ set files=FileSystemState Windows
for %%x in (%files%) do dot -Nfontname=simsun -T png -o %%x.png %%x.dot
```



3 changes: 2 additions & 1 deletion _posts/2012-02-11-State Machine and Event Processing.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@ layout: post
title: State Machine and Event Processing
original: https://blog.csdn.net/zedware/article/details/7251583
---

列了个提纲,一直没有写完。
1. Introduction
1.1 What's State Machine
Expand Down Expand Up @@ -31,4 +32,4 @@ Concurrent among states - inter-state concurrency
Wrapper for underlying APIs
Loops in a state
Loops among states


3 changes: 2 additions & 1 deletion _posts/2012-02-11-Weaving Relations for Cache Performance.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,11 +3,12 @@ layout: post
title: Weaving Relations for Cache Performance
original: https://blog.csdn.net/zedware/article/details/7251663
---

虽然这是篇较早的论文,但是对实现数据库存储的开发者,对要深入理解行式存储、列式存储、混合存储方式的技术人员来说还是很有参考价值的。严格的关系数据库理论并不要求应用程序开发者了解表的存储格式,但工程实践中并没有那么完美的好事。不同的存储格式对不同类型的应用或多或少存在一些实现/性能/使用上的影响。某些评论批评Oracle一直没有实现列式存储,所以对数据仓库的支持效率不够好。实际上甲骨文也是不得已,一个庞大的企业级软件哪像一个小软件那样可以随便改来改去啊。
Abstract
Relational database systems have traditionally optimzed for I/O performance and organized records sequentially on disk pages using the N-ary Storage Model (NSM) (a.k.a., slotted pages). Recent research, however, indicates that cache utilization and performance
is becoming increasingly important on modern platforms. In this paper, we first demonstrate that in-page data placement is the key to high cache performance and that NSM exhibits low cache utilization on modern platforms. Next, we propose a new data organization
model called PAX (Partition Attributes Across), that significantly improves cache performance by grouping together all values of each attribute within each page. Because PAX only affects layout inside the pages, it incurs no storage penalty and does not affect
I/O behavior. According to our experimental results, when compared to NSM (a) PAX exhibits superior cache and memory bandwidth utilization, saving at least 75% of NSM’s stall time due to data cache accesses, (b) range selection queries and updates on memory
resident relations execute 17-25% faster, and (c) TPC-H queries involving I/O execute 11-48% faster。


1 change: 1 addition & 0 deletions _posts/2012-02-12-有关Splay Tree的一点感想.md
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@ layout: post
title: 有关Splay Tree的一点感想
original: https://blog.csdn.net/zedware/article/details/7253327
---

Splay Tree是一种有趣的Binary Search Tree。Wikipedia说它的缺点为:

Perhaps the most significant disadvantage of splay trees is that the height of a splay tree can be linear. For example, this will be the case after accessing all *n* elements in non-decreasing order. Since the height of a tree corresponds to the worst-case access time, this means that the actual cost of an operation can be slow. However the [amortized](http://en.wikipedia.org/wiki/Amortized "Amortized") access cost of this worst case is logarithmic, O(log *n*). Also, the expected access cost can be reduced to O(log *n*) by using a randomized variant[[2]](http://en.wikipedia.org/wiki/Splay_tree#cite_note-1).
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@ layout: post
title: Disk Corruption以及它们对DBMS的影响
original: https://blog.csdn.net/zedware/article/details/7269074
---

众所周知,软硬件都有不靠谱的时候。可问题是一般人很难获取这些系统在实际使用中的故障率。作为这些产品的生厂商,一般都有一些方法来获取实际安装部署以及运行的情况,自然的,它们也掌握了很多实际的故障数据(不过它们未必会如实公开)。对于那些故障率不太高的情形,只有在安装部署的数量特别大时,才可能捕获到较大的故障出现次数。An Analysis of Data Corruption in the Storage Stack分析了存储系统中的Data Corruption:
An important threat to reliable storage of data is silent data corruption. In order to develop suitable protection mechanisms against data corruption, it is essential to understand its characteristics. In this paper, we present the first large-scale study
of data corruption. We analyze corruption instances recorded in production storage systems containing a total of 1.53 million disk drives, over a period of 41 months. We study three classes of corruption: checksum mismatches, identity discrepancies, and parity inconsistencies.
Expand Down
Loading

0 comments on commit d2c5795

Please sign in to comment.