Comments for John Jeffries's Blog

Comment on Overcoming ORA-01341: LOGMINER OUT-OF-MEMORY in Oracle Streams Downstream Capture by Lars Bo

Lars Bo — Mon, 21 Nov 2011 13:58:42 +0000

Thanks for sharing! Cheers Lars Bo

Comment on Using RMAN to restore a database to another server in an ASM environment by I'm glad with that

I'm glad with that — Thu, 29 Sep 2011 07:42:57 +0000

that is awesome!…

[...]Using RMAN to restore a database to another server in an ASM environment « John Jeffries's Blog[...]…

Comment on Re-instantiating Schema Replication following Database Incomplete Recovery by R. Hoekstra

R. Hoekstra — Fri, 05 Aug 2011 12:58:45 +0000

This is not a complete scenario to repair your downstream target database. If you do a PIT on your source database, what about the transactions that are already applied to your downstream target database after the time of the PIT?
On the Oracle site you find the document “Oracle Streams Replication Administrator’s Guide 10.2, 9 Managing Capture, Propagation, and Apply. How to configure Streams Real-Time Downstream Environment.” that contains a nearly complete scenario in the § called “Performing Point-in-Time Recovery on the source in a single-source environment”.

Comment on Packt Enterprise has just published the first ever GoldenGate book by John Jeffries

John Jeffries — Fri, 01 Apr 2011 10:21:51 +0000

Hi Kamran

Thank you for your kind comments and taking the time to review my book.

Regards
John

Comment on Packt Enterprise has just published the first ever GoldenGate book by Kamran Agayev A.

Kamran Agayev A. — Fri, 01 Apr 2011 10:06:05 +0000

Hi John. Hope you’re well

I’ve already started reading your book Will post my review on my blog and on amazon

My heartiest congratulations to you

Comment on About Me by John Jeffries

John Jeffries — Wed, 16 Mar 2011 08:57:36 +0000

Hi Jimmy

It looks like your apply process is the bottleneck. How many apply servers do you have configured?
Your apply process may benefit from an increase in parallelism.

The following example sets parallelism to 32:

BEGIN
dbms_apply_adm.set_parameter( apply_name => ‘TGT_SCHEMA_APPLY’,
parameter => ‘parallelism’,
VALUE => ’32′);
END;
/

Run the following query to display information about the apply servers for each apply process:

SELECT r.APPLY_NAME,
SUBSTR(s.PROGRAM,INSTR(S.PROGRAM,’(‘)+1,4) PROCESS_NAME,
r.STATE,
r.TOTAL_ASSIGNED,
r.TOTAL_MESSAGES_APPLIED
FROM V$STREAMS_APPLY_SERVER R, V$SESSION S
WHERE r.SID = s.SID AND
r.SERIAL# = s.SERIAL#
ORDER BY r.APPLY_NAME, r.SERVER_ID;

Regards
John

Comment on About Me by Jimmy

Jimmy — Mon, 14 Mar 2011 21:04:01 +0000

Hi John, I have 10g Streams set up one way and at the destination db, the apply process is running, but has 20,000 active transactions and 1 spilled — how do I go about getting all those active transactions to start decreasing? Currently, it’s continuing to grow — I guess this means they aren’t getting applied?

Thanks, Jimmy

Comment on Re-instantiating Schema Replication following Database Incomplete Recovery by oracle_dba_17

oracle_dba_17 — Tue, 26 Oct 2010 18:59:10 +0000

Nice job posting this document. Thank you!

Comment on Making permissions persistent on EMC PowerPath devices for CRS on Enterprise Linux by leslie

leslie — Thu, 01 Jul 2010 12:19:25 +0000

This method works for a small mount of devices as this is manual rename each powerpath device, it is unacceptable if we need to manually rename 50 disks on eight node cluster servers, is there any way we can use one node as primary server and all other nodes will copy powerpath pseudo name configuration from primary server so we can make consistent pseudo name accros cluster nodes. We want the process are not IO interruption.

Thanks

Comment on Making Oracle Streams Fly by Bous

Bous — Mon, 10 May 2010 11:40:50 +0000

I saw about 5 millions spill messages in queue on customer side , and it took them many many hours to clean the queue, Flow control is automatic in 10g and has a number of conditions that can trigger the capture process to pause for flow control.
1 memory pressure
2.Unseen messages.

If the apply is slow and the condition of flow control are not meet , the queue spilling will start spilling to AQ$_QUEUE_NAME_P.

regards,

Bous