Prev: How to find records with length greater than 17
Next: CHECKSUM to determine record changes
From: Dan.Cook on 21 Apr 2008 16:45
On Feb 28, 11:12 am, Teresa Masino <teresa.mas...(a)peninsula.org>
wrote:
> We have set up a couple of SQL Server 2005 systems and I have found
> that the format of the ERRORLOG files and the SQL Agent's log files
> are Unicode or some format that findstr cannot parse properly.  "find"
> parses them fine, but it doesn't have the capabilities that I need --
> specifically, I can't search for multiple strings in one search.
>
> I see the checkbox on the SQL Agent's for "Write OEM File", but it is
> grayed out so I am not able to try checking that.  I also don't know
> if that would affect the server's ERRORLOG file too or just the
> Agent's log file.
>
> So what am I missing?  What is everyone else doing who is used to
> having scripts to parse these files looking for strings that indicate
> problems?  Is there a server setting that will force it to go back to
> a plain ANSI text file format for log files?  Is that a bad thing to
> do?
>
> Thanks in advance for any insight,
> Teresa Masino

The TYPE command will convert unicode to ASCII. i.e.

TYPE \\coxsql2\E$\MSSQL.1\MSSQL\LOG\ERRORLOG > \\coxsql2\C$\ERRORLOG
\ErrorLog.txt
From: Dan.Cook on 21 Apr 2008 16:49
On Feb 28, 5:33 pm, Erland Sommarskog <esq...(a)sommarskog.se> wrote:
> Teresa Masino (teresa.mas...(a)peninsula.org) writes:
> > We have set up a couple of SQL Server 2005 systems and I have found
> > that the format of the ERRORLOG files and the SQL Agent's log files
> > are Unicode or some format that findstr cannot parse properly.  "find"
> > parses them fine, but it doesn't have the capabilities that I need --
> > specifically, I can't search for multiple strings in one search.
>
> > I see the checkbox on the SQL Agent's for "Write OEM File", but it is
> > grayed out so I am not able to try checking that.  I also don't know
> > if that would affect the server's ERRORLOG file too or just the
> > Agent's log file.
>
> > So what am I missing?  What is everyone else doing who is used to
> > having scripts to parse these files looking for strings that indicate
> > problems?  Is there a server setting that will force it to go back to
> > a plain ANSI text file format for log files?  Is that a bad thing to
> > do?
>
> I doubt there is a setting to force the SQL Server error log to be ANSI.
> Would be a lot problem to log object and database names with characters
> outside the ANSI set in that case.
>
> I guess the answer is to use Unicode-capable tools. I don't have any
> real suggestion for that. I search the errorlog fairly rarely, so most
> of the time I'm content of opening it in Textpad. (Which understands
> Unicode in so far it can handle the encoding, but it's not able to
> work with characters outside the ANSI range.)
>
> --
> Erland Sommarskog, SQL Server MVP, esq...(a)sommarskog.se
>
> Books Online for SQL Server 2005 athttp://www.microsoft.com/technet/prodtechnol/sql/2005/downloads/books...
> Books Online for SQL Server 2000 athttp://www.microsoft.com/sql/prodinfo/previousversions/books.mspx- Hide quoted text -
>
> - Show quoted text -

From a batch job the TYPE command will convert unicode to ASCII. i.e.

TYPE \\SQLSRVR1\E$\MSSQL.1\MSSQL\LOG\ERRORLOG > \\SQLSRVR1\C$\ERRORLOG
\ErrorLog.txt
From: rcamarda on 21 Apr 2008 18:31
check out LOGPARSER available from Microsoft. It can be used to parse
Active Directory and server event logs.
I use it to load data directly into a SQL table :)
HTH
Rob
 | 
Pages: 1
Prev: How to find records with length greater than 17
Next: CHECKSUM to determine record changes