Re: Current Redo got deleted
- From: "Mladen Gogala" <dmarc-noreply@xxxxxxxxxxxxx> (Redacted sender "mgogala@xxxxxxxxx" for DMARC)
- To: oracle-l@xxxxxxxxxxxxx
- Date: Sun, 22 Mar 2015 22:01:16 -0400
On 03/22/2015 09:34 PM, Don Seiler wrote:
This sounds good in theory, but less than 2 months ago our SAN had
what can only be described as a "freak out" which resulted in the
current (non-multiplexed) redo log being corrupted. It forced us to
failover to standby (and then use flashback database to turn the
original primary into a new standby). The SAN disks were mirrored
pairs and the redo disk group was external redundancy, FYI. We're
entertaining new storage vendors that also insist that we don't need
to multiplex redo logs with their magic RAID sauce but once bitten,
twice shy, as they say. I can't in good conscience leave us exposed to
that vulnerability again.
Don, what would have changed if the redo logs were duplexed? Where would
you dupliex them to? Another SAN? If SAN freaked out (your expression,
not quite sure what does it mean) and the duplexed redo logs were on
the same SAN, what would that duplex copy do for you? Yes, that's the
case for fail-over to standby. What protected you in this case was a
standby database, not duplex redo logs. I don't see that anything would
be different if you had another set of redo log files on the SAN that
has "freaked out". I am a big fan of standby databases, especially
active DG which can also be used as a reporting server.
I also have a question: do you normally run your production databases
with the flashback on? The cost of that can be quite severe. I would
have rebuilt the primary from backup.
--
Mladen Gogala
Oracle DBA
http://mgogala.freehostia.com
--
//www.freelists.org/webpage/oracle-l
Other related posts:
