垃圾Sybase 活该倒闭

Hangxiao

In order to make my comments more constructive. I just publish its usage here. So, guy thank you for you concentration.

dbcc monitor ( task, group, option )

Allows a more detailed monitoring of the SQL Server. [Hint: Run dbcctraceon(8399) first, to enable descriptive names for groups.]

      Task
        "clear" clear earlier gathered statistics
        "sample" gather fresh statistics
        "select" enable statistics to be displayed

      group "all" all groups' statistics are gathered/displayed
        group_name name of the group for which statistics are to be gathered/displayed

      option
        "on" enable monitoring
        "off" discontinue monitoring


dbcc traceon( tracenum [, tracenum ] ... )
Turn on specified trace flags.

      tracenum number of the trace flag to turn on

      Trace Flags
      NumberDescription
      200 "Before" image of query tree
      201 "After" image of query tree
      260 To prevent "done_in_proc" messages
      302 Information in index selection
      310 Information in join selection
      317 Complete information in join selection (voluminous)
      320 Turn off join order heuristic
      1204 Print out deadlock chains and victim
      1205 Print out stack traces of deadlocked pids. Use in conjuction with
      1204
      2512 Bypass syslogs when doing checkalloc
      1603 Use only standard unix calls for disk I/O
      1610 Disable packet batching
      3300 Display each log record as it is processed by recovery (voluminous)
      3402 Do a alloc upgrade during loaddb recovery
      3604 Send dbcc output to screen
      3605 Send dbcc output to errorlog
      3607 Do not recover any database, clear tempdb, or startup checkpoint
      3608 Recover master only. Do not clear tempdb, or startup checkpoint
      3609 Recover all databases. Do not clear tempdb, or startup checkpoint
      3620 Do not kill infected processes
      3701 Allows dropping and creating of system indexes (DANGEROUS!)
      4000 Crash server at next loop for recovery testing
      4012 Boot without starting up checkpoint process
      4013 Place a record into errorlog each time someone logs in to the server
      5101 Engine 0 will perform all disk and network I/O.
      5102 Prevents engine 0 from running any non-affinitied tasks
      8399 Enable descriptive names for groups (dbcc monitor)



A sample to use DBCC Monitor

use master
go
dbcc traceon(3604)
dbcc monitor ("clear", "all", "on")
waitfor delay "00:01:00"
dbcc monitor ("sample", "all", "on")
dbcc monitor ("select", "all", "on")
dbcc traceon(8399)
//so the monitor data is populated in the master.sysmonitors tables.
select field_name, group_name, value
  from sysmonitors
dbcc traceoff(8399)
go
dbcc traceoff(3604)
go

Hangxiao

A Sample To Watch How The Data Value Is Changed


A Sample To Watch How The Data Value Is Changed

/*Scenario 1 common usage in Sybase Agent*/
dbcc monitor("sample","all","on")
dbcc monitor("select","all","on")
dbcc monitor("clear","all","on")
dbcc traceon(8399)

/*wait for interval*/

    select GROUPNAME = group_name,
    FLDNAME    = field_name,
    FLDVALUE   = value
    from master..sysmonitors

/*clear the value for next time*/
dbcc monitor("clear","all","on");

/*****************************************************************************/
/*Scenario 2 for manually watching the metric change */

dbcc monitor("clear", "all", "on")
[wait]
dbcc monitor("sample", "all", "on")
dbcc monitor("select", "all", "on")
dbcc traceon(8399) -- to populate sysmonitors with the monitor counters, may not be needed in most recent versions
select * from sysmonitors
dbcc traceoff(8399)


/*Scenario 2 repeat the following the number in buffer_0 will be changed */

    select GROUPNAME = group_name,
    FLDNAME    = field_name,
    FLDVALUE   = value
    from master..sysmonitors
    where group_name = 'buffer_0'
    order by 1, 3 desc, 2

  
dbcc monitor("sample","all","off")
dbcc monitor("select","all","off")
dbcc monitor("clear","all","off")

/*Scenario 2 the following statement returns nothing, since it has been turned off*/
    select GROUPNAME = group_name,
    FLDNAME    = field_name,
    FLDVALUE   = value
    from master..sysmonitors
    where group_name = 'buffer_0'
    order by 1, 3 desc, 2

    select count(*) from sysmonitors

No data return


dbcc monitor("sample","all","on")
No data return

dbcc monitor("select","all","on")
    The data return  

dbcc monitor("clear","all","on")
The data is return

dbcc monitor("select","all","off")
No data return



Use the following you get an clear result

dbcc monitor("sample","all","on")

dbcc monitor("select","all","on")

dbcc monitor("clear","all","on")


   select GROUPNAME = group_name,
    FLDNAME    = field_name,
    FLDVALUE   = value
    from master..sysmonitors holdlock
    where group_name = 'buffer_0'
    order by 1, 3 desc, 2


Use the following, you get an un-clear result.

dbcc monitor("sample","all","on")

dbcc monitor("select","all","on")

//dbcc monitor("clear","all","on")


   select GROUPNAME = group_name,
    FLDNAME    = field_name,
    FLDVALUE   = value
    from master..sysmonitors holdlock
    where group_name = 'buffer_0'
    order by 1, 3 desc, 2


When there is a dbcc monitor("clear","all","on") statement, the counter number is clear.
and this will affect the number in sysmonitors.

〇〇

HEHE

donethat_cu

你的关于dbcc monitor的问题我实在没太明白，都是关于Sybase倒闭的说法倒是可以用“荒唐”或者“无知”来形容，看看过去10年的SY股票价格吧，附带和ORCL,IBM, MSFT做了个比较：
http://www.google.com/finance?ch ... =NYSE:SY&ntsp=0

hjb719

关于sybase资料不好找这点，我完全同意楼主观点，话说回来，能找到这篇帖子也是因为我在找dbcc monitor的相关资料才找到这里来

5个周

楼上的去SYBASE网站找过了么？？？？

andkylee

哎，不好找是真的。就看电子版的官方文档就行了~

zxp209

别啊，刚接触sybase，还不太深入，别打击新手信心啊！

andkylee

原帖由 zxp209 于 2010-3-19 18:34 发表
别啊，刚接触sybase，还不太深入，别打击新手信心啊！

不是打击，是事实。看你自己的抵抗力和毅力了~

linxun2004

同意，sybase的产品都未测试完成就发布了，SYBASE公司应该给我们第每个开发人员一份薪水，我们都盼着SYBASE公司倒闭呢，SYBASE因为中国的市场的特殊性（卖一套SYBASE给的回扣多），还有人民银行某些人的前瞻性，最后才是个人利益与SYBASE公司利益的共同结合，才造就的SYBASE公司的空前繁荣，只要中国市场的特殊性不改变，SYBASE还会一直发展状大。