What strategies can be followed to minimize downtime for patch installs, and reduce the likelihood of encountering problems?
We're currently on 184.108.40.206 and experiencing several database bugs. Heck, one's easy:
select 'asdf' from dual order by 1 desc; <-- causes process crash
All are fixed in 220.127.116.11 but having recently gone to 18.104.22.168 and through the motions of requesting downtime from our user communities, if there is something I can do to minimize downtime, that'd be great.
I've heard of people using logical standby: fail over from primary to secondary, patch primary, fail back to primary, patch secondary, something along those lines. Any other thoughts?
Secondly, what is a good strategy to evaluate the stability of a new patch release, and to prepare for the aftermath? I don't want to get burned again but is 22.214.171.124 going to cure our problems but introduce new ones? How to assess if a given patchset is "stable enough" ?
Originally posted by OracleDoc errr ummm apply it to dev first to see how long it's going to take and to see if it fixes your problem? That is if your Dev is in sync with prod...
Hi! Thank you for your reply. Unfortunately that is the easy part, testing to see if a bug is fixed that is. The thing I'm much more interested in is determining the degree to which the patch will cause problems of its own. Just installing in development may not reveal these issues; our development is fully in sync with production but with two hundred applications and half as many IT staff, we're not likely to "stumble" onto the esoteric bugs, like the ones biting us in 126.96.36.199 for example.
I know there is no clear answer to this question. I'm just wondering what other DBAs do to establish a comfort level with patches.
Originally posted by swiego I'm just wondering what other DBAs do to establish a comfort level with patches.
If I was constantly getting hit with a bug that is fixed in a higher level patch, I would tend to apply the patch so I don't get hit with that error. Will there be new bugs in the patch? Sure. What software doesn't have bugs? Besides, you can always go back to 188.8.131.52 if things really suck, right?
When applying patches, I prefer to install Oracle in a seperate $OH. I've outlined the reasons before of why I like to do that. By using multiple $OHs, you could be back on 184.108.40.206 in about 30 minutes.