Yesterday morning, I ran across this blog entry from Bobby Durrett’s blog about a typical day in his life as a DBA. So I figured I would have some similar fun and document my workday. So here is how my day went yesterday:
7:30 – Got to work. Checked email to make sure there was nothing that needed my immediate attention.
7:35 – Fired up Firefox and clicked on Morning Coffee. I reviewed new postings to the many blogs that I read and other technical sites to keep up on the latest database news.
7:50 – Answered questions on http://communities.oracle.com
8:15 – Received an email about an upcoming audit for Microsoft licensing. Needed to generate an account and provide access to our SCCM database in SQL Server. I also talked with a manager about this activity.
8:25 – Code reviews. Looked at all new code checked into source control since yesterday. I only concentrate on code changes that affect the database. Will continue code reviews periodically throughout the day otherwise I get behind.
9:10 – Talked with management about retiring some old SQL Server databases since the application that is using that has been deprecated.
9:20 – Worked on a connection issue users reported with SQL Server. Boy I’ve been working on a lot of SQL server stuff this morning so far. There should be no reason the users cannot connect. My test user can connect just fine which is in the same domain group as the users experiencing the problem. Need to get ahold of an end user to verify some results so this is on hold for a while.
9:35 – Testing switchover to standby database. Finally on to some Oracle stuff. When doing a switchover in production, got a timeout and switchover did not complete. Found a Note on Metalink that it could be related to an AFTER LOGON trigger we have in place. I am testing out this theory on a testbed. Seems early on in my testing, I broke my testbed. This is why we don’t just go and do these sorts of things in production without adequate testing upfront. As I am working through this issue, my brain is telling me that either I need to document the issues and steps I am taking because I will need this later or this would be a good test case to blog about…or both. So this work may take a bit longer as I document things. Finally got my primary and standby back up and running and I’ll tackle this a bit later after I catch up on a few things.
10:30 – Another DBA on the team just built two new development databases, clones of our production database. I checked out the new environments. One of the environments had a bad listener.ora so fixed that issue. I updated our documentation on these new environments, chatted with the other DBA for a bit and then started the job of deploying the agent to add these new systems to Enterprise Manager 12c. Finally, I let the other DBA on the team know these were ready.
11:00 – Verified everything is ready to go for tonight’s production maintence window.
11:05 – More code reviews
11:10 – Touched base with a DBA on the team about their learning track. He is new to being a DBA and needs mentoring. Among other things, we went over the new database environments and also how to add them to EM12c.
11:40 – Back to the My Oracle Support Communities to answer a few questions.
11:50 – I researched “Comparing” in SQL Developer that was given as a tip on ThatJeffSmith.com as I need to increase my skillset as well.
12:00 – Lunch. At lunch, I had a discussion with one of the software developers related to SQL%NOTFOUND in PL/SQL blocks.
12:45 – More code reviews
12:55 – Back to working on my standby switchover testing from earlier today. Disabled the AFTER LOGON trigger and the switchover worked. Disabling the trigger stopped the switchover and re-enabling the trigger allowed the switchover, so my testing is confirming that the trigger halted my previous switchover in production.
1:35 – I talked with the Software Development manager about a learning session/tutorial for the development staff on how to use Explain Plans. We also talked about ideas for other future learning sessions.
1:50 – Back to the My Oracle Support Communities to answer a few questions.
1:55 – Spent some time talking with an application developer about the best method of emailing reports generated in SQL Server.
2:15 – I worked on adjusting metric thresholds in Enterprise Manager and clearing down production alerts for issues that have been fixed. I am also trying to set up EM to send alerts to an Exchange email distribution list. I discovered that my SMTP relay will not allow the distribution list so I sent the issue to my server guys.
2:45 – More code reviews
2:55 – Production issue. Normally don’t get too many of these. It looks like a network issue cause connectivity problems with the databases. The instances were all up and running. But Enterprise Manager lost contact with one of the instances and is reporting it to be down when it is up and running. I had to restart the EM agent to fix the issue.
3:20 Got word back from the server team that the SMTP relay will now handle my distribution list. I configured all 3 EM environments to alert to the DL.
3:45 – More code reviews
3:50 – One Oracle production database on Windows has its C: drive too full. I made a production change to move the DIAGNOSTIC_DEST to the D: drive which has much more space. Since this is a production change, made sure to update the change tracking system.
4:00 – I read a nice Oracle white paper that dealt with Oracle 12c’s optimizer stats.
4:10 – I got interrupted on reading that white paper because an end user wanted to know why their session got disconnected from the database. I determined this was due to the earlier production issues related to the network. Now back to that white paper.
4:20 – Never got through the white paper. It will have to wait until tomorrow. I spent the last few minutes doing some last minute code reviews (a light day for code reviews today) and one last visit to the My Oracle Support Communities. My end users never got back to me about the connection problems to SQL Server so it must not be that important.
Tonight at 9pm, I will be performing database maintenance. And I have a Data Change Request that came in today which will probably run about 6pm. So even though I’m heading out the door, the day isn’t over.
All in all, this was a typical day for me. The only exception was that I did not tackle any performance problems today. And I did seem to work on SQL Server more than I normally do.