GPUTeraSort: High Performance Graphics Coprocessor Sorting for Large Database Management

GPUTeraSort: High Performance Graphics Coprocessor Sorting for Large Database Management
GPUTeraSort is a novel external sorting algorithm using graphics processors (GPUs) on large databases composed of billions of records and wide keys. It uses the data parallelism within a GPU along with task parallelism by scheduling memory-intensive and compute-intensive tasks on the GPU. This sorting architecture leverages multiple memory interfaces on the same PC . using both the high-bandwidth GPU memory interface along with the general CPU main memory interface. GPUTeraSort is structured as a task pipeline: read disk, build keys, sort using the GPU, generate runs, write disk, and then in the second phase, read, merge, write. As a result of these design decisions, it achieves higher memory bandwidth than purely CPU-based algorithms running on commodity PCs. It takes into account the limited communication bandwidth between the CPU and the GPU, and reduces the data communication between the two processors. It also pipelines disk transfers and achieves near-peak I/O performance.

Advertisements

Lufthansa Amateur Radio Club Frankfurt

Lufthansa Amateur Radio Club Frankfurt
im Lufthansa Sportverein Frankfurt e.V. (LSV) Sparte Amateurfunk.

Die Mitglieder der Sparte sind Funkamateuren aus verschiedenen Bereichen des Lufthansa Konzerns und OM des DARC OV F44 Rhein-Main.

DLØDLH

ist unsere Clubstation in der Nähe des Flughafens Frankfurt Rhein-Main. Die Ausrüstung erlaubt Funkbetrieb auf Kurzwelle sowie auf VHF/UHF in SSB, CW und digitalen Modes.

Ordinal Technology – Nsort Home Page

Ordinal Technology – Nsort Home Page
Nsort has a full set of features to meet your sorting needs. Nsort can do the following:

Sort large data sets quickly.

Merge already-sorted input files.

Use multiple processors and disks in parallel.

Read sort input from multiple input files.

Access large files with 64-bit file offsets.

Handle a variety of record types:

fixed-length

delimited (lines of text)

Handle a variety of key types:

character

decimal character

binary integer

IEEE floating point

month

Sort records and keys up to 64 Kbytes wide.

Sort on an unlimited number of ascending or descending keys.

Delete duplicate-keyed records.

Add, drop, and reorder the fields in the sort records.

Include or omit records conditionally.

Subtotal fields by key value.

Produce multiple output files, each with its own selection criteria.

Unix sort program-type key specifications for records with delimited fields.

ODA – Tool for Oracle objects Dependency Analysing

ODA – Tool for Oracle objects Dependency Analysing
The ODA tool has been written to analyze database dependencies between database objects such as procedures, tables, views etc. The tool can also be used to analyze database dependencies between objects in Forms/Report PL/SQL code and block properties and the database objects dependencies also. The ODA find object usage in the flat files(Unix scripts,C,XML) ,Informatica workflows and Word files.

HyperSQL

HyperSQL
HyperSQL is like a doxygen plus Javadoc for SQL, hypermapping SQL views, packages, procedures, and functions to HTML source code listings and showing all code locations where these are used. The internal “where used” functionality also scans C++ and Java source files.

AskTom:

AskTom:
We first need to break up the table into small pieces. We can do this by some numeric range – useful for tables that use a SEQUENCE to populate their primary key, by any arbitrary SQL you want to code, or by ROWID ranges. We’ll use the ROWID range, I find that to simply be the most efficient – it creates non-overlapping ranges of the table (contention free) and doesn’t require querying the table to decide the ranges, it just uses the data dictionary. So, we’ll make the following API calls:

big_table%ORA11GR2> begin

2 dbms_parallel_execute.create_task(‘PROCESS BIG TABLE’);

3 dbms_parallel_execute.create_chunks_by_rowid

4 ( task_name => ‘PROCESS BIG TABLE’,

5 table_owner => user,

6 table_name => ‘BIG_TABLE’,

7 by_row => false,

8 chunk_size => 10000 );

9 end;

10 /

PL/SQL procedure successfully completed.

We started by creating a named task – ‘PROCESS BIG TABLE’ in this case.

Roche migriert 90.000 Arbeitsplätze auf Google Apps

Roche migriert 90.000 Arbeitsplätze auf Google Apps
Roche, ein Unternehmen aus dem Bereich Healthcare mit weltweit über 90.000 Mitarbeitern, stellt jetzt komplett auf die Cloud-basierte Office-Alternative von Google um.

“Die Art und Weise, wie unsere Angestellten kommunizieren und kollaborieren ist sehr unterschiedlich und die Angestellten sind über 140 Länder verteilt”, erklärt Hippe. Dabei sei die Architektur der Google Apps sehr ansprechend. Hippe hofft nun, dass die Google-Lösung “das Unternehmen näher zusammenbringt”. Jetzt könnten Mitarbeiter einfach über ein Control-Panel zusammenarbeiten und es müssten dafür keine komplexe Infrastrukturen in den Rechenzentren eingerichtet werden.

Auch ohne ein VPN könnten die Mitarbeiter dann von überall her auf ihre Daten zugreifen. Damit könnten Angestellte zum Beispiel ohne Probleme auch von zu Hause aus arbeiten.

Nach der spanischen Bank BBVA, die bereits im Januar angekündigt hatte, 110.000 Arbeitsplätze auf Google Apps migrieren zu wollen, ist Roche nun der zweitgrößte Google Apps Anwe