[Info-Ingres] log writes vs dual log writes

Programming and Web Development Forums - DATABASE - Issues relating to INGRES products.

Skip to content

[Info-Ingres] log writes vs dual log writes

DATABASE
Post a reply

    Sponsored Links

    Next

  • 1. long byte cloumn
    Hi, I'm trying to use a LONG BYTE column type in Ingres II width cariboulake jdbc driver. I'm using setBinaryStream to insert the stream (file) into the column and resultSet.getBinaryStream to retreive it. Everything seems to work well when I insert little files (around 75k or less) but when the file is bigger, it seems to insert ok, but I can't retreive it. Whe I call getBinaryStream, I get an empty stream. any idea? thanx!
  • 2. Sizing - Number of disks for optimal write speed.
    Hi. I'm looking for case studies/advice regarding typical numbers of disks to use for a fairly large database [*]. We want to speed-up disk writes [singelton and bulk insert or update statements] so I am advising use of more disks [typically we use 5 data locations and ask that these be separate disks but in practice our users have RAID5 SANs for which the setup/details are opaque]. Does anyone have metrics on using more disks? e.g. point of diminishing returns. Alternatively, any other suggestions for speeding up writes without drastic code changes. I sometimes see 3-4 secs for a singleton insert more than once during the run of a single batch program. [*] The Db sizes range from 10-50Gb. There are several 100 tables but the majority of the data is held in just two. These tables grow by 100's of K's of rows each day. Unfortunately the keys comprise multiple columns and it is possible to have 1000's of rows with identical keys baring the right-most column. The latter is an ascending numeric identifier. We run Ingres 2.6 on Solaris. ta Philb
  • 3. ingres + veritas volume manager
    Anyone ever use ingres on a solaris machine with veritas volume manager? If so, any comments on experience - good or bad... we're looking at replicating to a remote server with the above. Gareth Williams English:- This message is confidential. If you are not the intended recipient of the message then please notify the sender immediately. Any of the statements or comments made above should be regarded as personal and not necessarily those of Bro Morgannwg NHS Trust, any constituent part or connected body. Cymraeg:- Mae'r neges hon yn gyfrinachol. Os nad chi yw'r derbynnydd y bwriedid y neges ar ei gyfer, byddwch mor garedig ?rhoi gwybod i'r anfonydd yn ddi-oed. Dylid ystyried unrhyw ddatganiadau neu sylwadau a wneir uchod yn rhai personol, ac nid o angenrhaid yn rhai o eiddo Ymddiriedolaeth GIG Bro Morgannwg, nac unrhyw ran gyfansoddol ohoni na chorff cysylltiedig.
  • 4. Perl DBD::Ingres question
    I'm not sure if this is a perl issue/question or Ingres issue/question but, ... I have a perl script that runs on a server in the PST timezone that connects to a remote dB through a vnode connection. Within the perl script I change the timezone with $ENV{TZ} = "EST5EDT", for example, and issue a query that returns a date. I get a date in the EST timezone, which is what I would expect. If I then do $ENV{TZ} = "GMT" and issue the same query against the same remote dB, the date still comes back in the EST timezone. I've tried disconnecting, changing tz, and reconnecting (all in the same script) but that doesn't make a difference either. Is this an issue with how perl handles environment variables or some caching of stuff by the dbms server? I thought I had remembered there was a II_TIMEZONE variable that could come into play, but perusing the manuals I couldn't find it. I have other scripts that change II_SYSTEM, for example, on the fly to point to various installations based on underlying character sets which seem to work OK. Thanks Dennis

[Info-Ingres] log writes vs dual log writes

Postby Martin Bowes » Fri, 24 Oct 2008 18:30:04 GMT

his is a multi-part message in MIME format.

Hi All,



We've restarted an installation after a database went inconsistent, the
database was recovered from a checkpoint. The system was restarted OK.



Since then ... without any sign of a problem in the errlog... I'm seeing
a major disparity in the log writes and dual log writes as reported in
logstat.



write complete 1 dual write complete 554524

Status: ONLINE,ECPDONE,JSWITCHDONE

Active Log(s): II_LOG_FILE,II_DUAL_LOG



The complete lack of activity on the primary log since startup is
reinforced by a simple OS check.

flourish.ingres> cd /dbsystem/II/ingres/log/

flourish.ingres> ls -l

total 2001960

-rw------- 1 ingres ingres 2048000000 Oct 14 11:16 ingres_log.l01

flourish.ingres> cd /dbdual/II/ingres/log/

flourish.ingres> ls -l

total 2001960

-rw------- 1 ingres ingres 2048000000 Oct 23 10:18 dual_log.l01



Anyone seen anything like this before?



I'm going to shutdown and rebuild my primary log file and see if that
fixes the problem.



Martin Bowes


<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40">

<head>
<meta http-equiv=Content-Type content="text/html; charset=us-ascii">
<meta name=Generator content="Microsoft Word 12 (filtered medium)">
<style>
<!--
/* Font Definitions */
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0cm;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri","sans-serif";}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:purple;
text-decoration:underline;}
span.EmailStyle17
{mso-style-type:personal-compose;
font-family:"Calibri","sans-serif";
color:windowtext;}
..MsoChpDefault
{mso-style-type:export-only;}
@page Section1
{size:612.0pt 792.0pt;
margin:72.0pt 72.0pt 72.0pt 72.0pt;}
div.Section1
{page:Section1;}
-->
</style>
<!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]-->
</head>

<body lang=EN-GB link=blue vlink=purple>

<div class=Section1>

<p class=MsoNormal>Hi All,<o:p></o:p></p>

<p class=MsoNormal><o:p> </o:p></p>

<p class=MsoNormal>We've restarted an installation after a database went
inconsistent, the database was recovered from a checkpoint. The system was
restarted OK.<o:p></o:p></p>

<p class=MsoNormal><o:p> </o:p></p>

<p class=MsoNormal>Since then … without any sign of a problem in the
errlog… I'm seeing a major disparity in the log writes and dual log
writes as reported in logstat.  <o:p></o:p></p>

<p class=MsoNormal><o:p> </o:p></p>

<p class=MsoNormal>
Top

Re: [Info-Ingres] log writes vs dual log writes

Postby Martin Bowes » Fri, 24 Oct 2008 19:09:01 GMT

his is a multi-part message in MIME format.

It did.



Interestingly, the shutdown was clean and yet cbf still said there were
uncommitted transaction in the log file.



Marty



From: XXXX@XXXXX.COM
[mailto: XXXX@XXXXX.COM ] On Behalf Of
Martin Bowes
Sent: 23 October 2008 10:30
To: XXXX@XXXXX.COM
Subject: [Info-Ingres] log writes vs dual log writes



Hi All,



We've restarted an installation after a database went inconsistent, the
database was recovered from a checkpoint. The system was restarted OK.



Since then ... without any sign of a problem in the errlog... I'm seeing
a major disparity in the log writes and dual log writes as reported in
logstat.



write complete 1 dual write complete 554524

Status: ONLINE,ECPDONE,JSWITCHDONE

Active Log(s): II_LOG_FILE,II_DUAL_LOG



The complete lack of activity on the primary log since startup is
reinforced by a simple OS check.

flourish.ingres> cd /dbsystem/II/ingres/log/

flourish.ingres> ls -l

total 2001960

-rw------- 1 ingres ingres 2048000000 Oct 14 11:16 ingres_log.l01

flourish.ingres> cd /dbdual/II/ingres/log/

flourish.ingres> ls -l

total 2001960

-rw------- 1 ingres ingres 2048000000 Oct 23 10:18 dual_log.l01



Anyone seen anything like this before?



I'm going to shutdown and rebuild my primary log file and see if that
fixes the problem.



Martin Bowes


<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40">

<head>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=us-ascii">
<meta name=Generator content="Microsoft Word 12 (filtered medium)">
<style>
<!--
/* Font Definitions */
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
@font-face
{font-family:Tahoma;
panose-1:2 11 6 4 3 5 4 4 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0cm;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri","sans-serif";}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:purple;
text-decoration:underline;}
span.EmailStyle17
{mso-style-type:personal;
font-family:"Calibri","sans-serif";
color:windowtext;}
span.EmailStyle18
{mso-style-type:personal-reply;
font-family:"Calibri","sans-serif";
color:#1F497D;}
..MsoChpDefault
{mso-style-type:export-only;
font-size:10.0pt;}
@page Section1
{size:612.0pt 792.0pt;
margin:72.0pt 72.0pt 72.0pt 72.0pt;}
div.Section1
{page:Section1;}
-->
</style>
<!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]-->
</head>

<body lang=EN-GB link=blue vlink=purple>

<div class=Section1>

<p class=MsoNormal><span style='color:#1F497D'>It did.<o:p></o:p></span></p>

<p class=MsoNormal><span style='color:#1F497D'><o:p> </o:p&
Top

Similar Threads:

1.[Info-Ingres] logstat write vs dual write 

Hi Everyone,

Does anyone have a strong feeling on how much attention I should be 
paying to discrepancies in the Log Write and Dual Log Write figures in
logstat?

I've always checked that these numbers are much the same - seems 
like a reasonable thing to do. Fact is I have a monitor program that
sounds an alarm if they are more than 10 writes apart.

On most of the systems I've got, I see these numbers never being more 
than
a few writes out, and that over time they tend to always agree to a very
small percentage. So my monitor very rarely pings at me.

But on some hosts, the writes slowly drift further apart - generally as a
result of load, and never seem to come back into agreement. The 
percentage
difference is small, but troubling.

eg.
On the 5th Jan at 10:51:
    write complete         868522        dual write complete    868523

Now:
    write complete        2587889        dual write complete   2587759

The system drifted in response to heavy usage of the log file on the 5th
Jan. I've checked in the OS system logs and there is no sign of a 
problem
on the disk holding the dual log.

Martin Bowes
Random Duckman Quote #26:
Cornfed - Duckman had been reduced to the state of a gibbering, 
mindless idiot. In many ways this was an improvement - but his pro 
country music attitude was starting to worry me.

2.[Info-Ingres] Emma McGrattan : Men Write Code from Mars, Women Write More Helpful Code from Venus

3.How can i write a log query sequence? 

Hi at all!
I need to write in a text file all the query that Informix applied in
a database.
I can do that?...and how? 

Tnx at all!

4.ontape could not write log tape error is coming 

Hi,

could not write log tape error is coming while taking ontape
-c.Informix version is 7.3
tar command is working fine.In config file  device entry is  /dev/rct0.

First the logical log backup is directed to harddisk.After  DDS3-dat
drive has installed.

Guideme to restore the problem.

tks
palani

5.How to force logical log write to tape? 

Presently, our Informix database logical logs is configured to write
to our DLT tape library.

I had this problem when changing tape in DLT Tape Library in Legato
Networker.
When a mounted tape in the tape library was being temporary ejected,
and before it was fully ejected, and if there is any Informix log
write happened. Networker would marked the tape in transit missing,
and prompt for next tape, even if the "tape-in-transit" was under
utilised.

In order to prevent this from happening, I need to flush all the logs
to tape prior to the intended tape ejecting. This would allow me more
time to complete the ejecting cycle.

There is no Informix database administrator in our workplace.
Anyone could tell me the syntax to do so?

6. XLogWrite: write request 0/53A4000 is past end of log

7. XLogWrite: write request 0/53A4000 is past end of log 0/53A4000

8. [Info-ingres] 3rd party tools for writing/managing Ingres rules and procedures (amongst other things)

Post a reply

Return to DATABASE

 

Who is online

Users browsing this forum: No registered users and 30 guest

Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group