Adding Hypertable API to DBI

As of now C++ is the only language that will interface with Hypertable, which is unfortunate, because I have large complex Perl scripts that would take forever to convert to C++ and my productivity would drop 90%. I am considering writing a Perl module for Hypertable, however, DBI is really the

RE: perl dbi and RaiseError handling - is this a situations which won't raise an error?

Hi Larry, More on Douglas, the execution of queries such DELETE, INSERT or UPDATE returns the number of affected rows. For no row affected the DBI module will return 'OEO'. Meaning 0 multiplied by 10 exp 0. So in Boolean context it will be true and in numerical context it will be 0. If an error

Re: What is the meaning of the rows value after a select?

On Mon, Dec 1, 2008 at 8:52 PM, Douglas Wilson [EMAIL PROTECTED] wrote: On Mon, Dec 1, 2008 at 7:54 AM, Larry W. Virden [EMAIL PROTECTED] wrote: Also, since RaiseError is set, the or die... part of this will never get executed either: $getMatchRec-execute() or die Couldn't fetch records

Protecting a mod perl 1.3 site from slow MySql processes

Hello, I am working on a site that has a modperl (mod_perl 1.3, perl 5.8) handler doing some authentication work via DBI::MySql. It connects to a MySQL server over the network. Recently during some maintenance on the MySql server, the entire site was hung up, because the MySQL server was up, but

Re: Adding Hypertable API to DBI

Some searching leads me to believe that Thrift is being used to implement a broker interface to Hypertable. http://qconsf.com/sf2008/file?path=/qcon-sanfran-2008/slides//DougJudd_Hypertable.pdf You might also want to consider http://thrudb.org/ Tim. On Mon, Dec 01, 2008 at 06:28:02PM -0800,

DBI + ithreads performances

Dear users, I need to fetch data from two different database (A and B). I wrote a sequential code fetching from A and, after the fetch is finished, from B. Then I tried to use threads: I splitted the fetches into 2 threads: each one makes its own connection to A or B and fetches his data.

Re: DBI + ithreads performances

On Tue, 2008-12-02 at 18:01 +0100, Davide Sacchetti wrote: I need to fetch data from two different database (A and B). I wrote a sequential code fetching from A and, after the fetch is finished, from B. Then I tried to use threads: I splitted the fetches into 2 threads: each one makes its

Re: Protecting a mod perl 1.3 site from slow MySql processes

alarm() definitely works in mod_perl. For timeouts in DBI see: http://search.cpan.org/~timb/DBI-1.607/DBI.pm#Signal_Handling_and_Canceling_Operations Hendrik Am Di, 2.12.2008, 16:22, schrieb April Papajohn (Blumenstiel): Hello, I am working on a site that has a modperl (mod_perl 1.3, perl

Re: What is the meaning of the rows value after a select?

On Tue, Dec 2, 2008 at 4:27 AM, Larry W. Virden [EMAIL PROTECTED] wrote: With the DBI RaiseError condition set, all DBI related statements are going to raise an exception, so none of the or die statements are necessary ... they won't hurt (or help), but their presence may deceive the novice

Re: Protecting a mod perl 1.3 site from slow MySql processes

On Tue, Dec 2, 2008 at 2:19 PM, Hendrik Schumacher [EMAIL PROTECTED] wrote: alarm() definitely works in mod_perl. Then I wonder why my test code is successfully interrupting the DBI-connect method from the command line, but not when running in modperl 1.3 (the versions of perl are the same).

Re: Protecting a mod perl 1.3 site from slow MySql processes

You also might want to look at the way DBIx::HA implements its timeout checks and overall analysis of the DB's responsiveness. The source code is very small so shouldn't be difficult to follow. H. On Dec 2, 2008, at 9:19 PM, Hendrik Schumacher wrote: alarm() definitely works in mod_perl.