18 messages in org.apache.logging.log4j-userRe: [log4j] FW: Database Logging Perf...| From | Sent On | Attachments |
|---|---|---|
| Vernier, Dominique (DVernier) | 14 Jun 2004 05:05 | |
| Ceki Gülcü | 14 Jun 2004 05:14 | |
| Vernier, Dominique (DVernier) | 14 Jun 2004 05:16 | |
| Ceki Gülcü | 14 Jun 2004 05:30 | |
| pooja | 14 Jun 2004 06:19 | |
| Vernier, Dominique (DVernier) | 14 Jun 2004 11:18 | |
| Ceki Gülcü | 14 Jun 2004 11:26 | |
| Vernier, Dominique (DVernier) | 14 Jun 2004 11:52 | |
| mrut...@westpac.com.au | 15 Jun 2004 16:24 | |
| mrut...@westpac.com.au | 15 Jun 2004 16:24 | |
| pooja | 18 Jun 2004 00:31 | |
| Andreas Guther | 18 Jun 2004 05:27 | |
| Thomas Svensen | 18 Jun 2004 05:37 | |
| Ceki Gülcü | 18 Jun 2004 06:02 | |
| Nima | 18 Jun 2004 09:50 | |
| Andreas Guther | 22 Jun 2004 19:21 | |
| Ceki Gülcü | 23 Jun 2004 07:41 | |
| Andreas Guther | 24 Jun 2004 04:32 |
| Subject: | Re: [log4j] FW: Database Logging Performance Numbers |
|---|---|
| From: | Ceki Gülcü (ce...@qos.ch) |
| Date: | 06/18/2004 06:02:00 AM |
| List: | org.apache.logging.log4j-user |
My tests with DBAppender show that using pooled connections, it takes about 10 milliseconds to insert an event into a MySQL DB (of type *INNODB*) running locally. It still takes approx. 10 millis logging to a PostgreSQL DB running on a relatively old and slow Linux server (with pooled connections).
DBAppender is completely DB agnostic. It will run on all RDBMMSs. An optimized version of DBAppender (but which is MySQL-only) logs at 500 *microseconds* per event.
HTH,
At 02:27 PM 6/18/2004, you wrote:
I am also interested in that kind of information. I also started recently looking into logging into tables and would be interested in some performance results and experience reports in high volume environments.
Andreas
pooja wrote:
Hi All I haven't received a response on this issue. Can someone at Log4j answer it for me? Thanx --- Pooja.
-----Original Message----- From: pooja [mailto:poo...@induslogic.com] Sent: Monday, June 14, 2004 6:50 PM To: 'Log4J Users List' Subject: Database Logging Performance Numbers
Hello I am looking for some performance numbers for Database Logging. I want to be able to analyze/query my log records at a later date in multiple of ways. Hence Logging into database seems to be the perfect choice. But the system is a transactional system and it will be handling more than a million transactions a day. Hence making it necessary, that I first prove that Logging into Database will not hamper performance. Can someone provide some published numbers?
I will be fine with test cases as well; I'll run them against the current file logging system and get comparative figures. Currently, I am writing some crude test cases of my own. At the same time, if neone knows of some application (which handles so many transactions) already writing logs into database & functioning, this reference will suffice my job. Thanx --- Pooja.
-- Ceki Gülcü
For log4j documentation consider "The complete log4j manual" ISBN: 2970036908 http://www.qos.ch/shop/products/clm_t.jsp