Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: redo log sync time

Re: redo log sync time

From: Brian Peasland <peasland_at_edcmail.cr.usgs.gov>
Date: Tue, 9 Jan 2001 15:54:57 GMT
Message-ID: <3A5B3451.D5A9B2F0@edcmail.cr.usgs.gov>

Andreas,

Just in case you didn't know, or in case others need to know: The redo log sync time is a normal event that occurs when a COMMIT takes place. Before a transaction is committed, the redo information is written to the online redo logs. The redo log sync time is the amount of time that the transaction waits for the redo information to be written to the online redo logs. There are multiple reasons why you could see a high *average* wait time for this event.

  1. How many transactions are committing at this time? If you have a high number of commits, then your transactions may have to wait for others to finish writing to the online redo logs.
  2. Is there other disk I/O on that same device? You might be waiting for disk head seek-time or rotational-delay if there is other disk activity on the same devices as your online redo logs. If possible, isolate the logs to their own, dedicated disk devices.
  3. Is there other hardware related disk processes? For instance, are you using hardware caching or hardware logging for recoverability? Are you using raw files or an O/S file system? Raw files can speed this up.
  4. What are the manufacturer's specs on the speed of their disk devices? Are they "fast" enough for you? You may have to upgrade to faster disk devices.

HTH,
Brian  

Andreas Goldammer wrote:
>
> A few weeks ago I found a high average redo log sync time (40 ms)
> for my AIX database. I looked at the documentation and
> everything I found was tuning the I/O subsystem (SSA) or
> reduce the number of commits (impossible, we are using SAP R/3).
>
> Two weeks ago I moved the database to new hardware
> (new fast box, new fast adapters, new fast disks) and
> removed the AIX mirror from the online redo log filesystems
> to avoid the MWC impact (now I use ORACLE to mirror the log)
>
> The average redo log sync time remains steady, whereas
> other file access counters have decreased
>
> Any other ideas beside I/O and commits ?
>
> Thanks in advance
> Andreas
> mailto: agoldammer_at_viega.de
 

-- 
========================================
Brian Peasland
Raytheons Systems at
  USGS EROS Data Center
These opinions are my own and do not
necessarily reflect the opinions of my 
company!
========================================
Received on Tue Jan 09 2001 - 09:54:57 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US