A recent MOS document discusses, albeit briefly, an issue with AWR reports that’s been going on since 10.2.0, namely that the elapsed time numbers for queries executed in parallel are considerably greater than the actual elapsed clock time. Let’s look at why that is and what can be done about it.
AWR reports were a tremendous improvement over Statspack reports, primarily due to the depth and breadth of the sample data collected. A (more...)
Editor’s note: In January of 2014, our team held a wearables summit of sorts, test-driving five popular watches, fitness bands and head-mounted displays to collect experiential evidence of each form factor, initial experience, device software and ecosystem and development capabilities.
I conducted customer feedback sessions with users who fit the “c-level executive” user profile, to collect feedback on some of our new interactive data visualizations. Unfortunately, I can’t share any of these design concepts just yet, but I can share a bunch of pics of Noel, who gave several (more...)
In a forum I contribute to the following question was asked:
Can adaptive cursor sharing (ACS) depend on execution order ?
The issue described a relatively basic query that changed execution plans apparently due to the order the query statements were run based on bind variable values. It’s an interesting issue that testing has verified. The tests are reproduced below, in abbreviated form, so let’s look at what was executed and what execution plans were used.
It would appear that some DBAs are still using the optimizer_index_cost_adj parameter to make index access paths more ‘desirable’ to the optimizer. In decades past this might have been a good strategy however with the improvement in statistics gathering in recent relesaes of Oracle this might not be the case. Let’s look at an example to see why this might do more ‘harm’ than good.
The optimizer_index_cost_adj parameter was first provided in Oracle 9i as (more...)
An interesting ‘problem’ surfaced a while ago, one where a user with zero quota on every tablespace could successfully create tables. Of course once it was time to insert data the inserts failed, but this was confusing the user creating the tables. The ‘problem’ stems from enabling deferred segment creation in the database. Let’s see how that can create a confusing situation,
Deferred segment creation allows tables and indexes to be created without physical segments (more...)
As I help a peer with a SQL Tuning engagement, I face the frequent case of: “We do not want to gather CBO schema object statistics because we don’t want our Execution Plans to change”. Well, the bad news is that: not gathering stats only gives you a false sense of safety because your Execution Plans can change anyways. The reason has to do with Predicates referencing values out of range. Typical cases include range (more...)
This week one of our developers highlighted the fact that the DBMS_SCHEDULER on one of our 220.127.116.11 databases was running jobs an hour later than scheduled until just recently, when the clocks went back (26/10/2014) for British Summer Time (BST) …and now they’re running on-time again! I’d not seen this before, and at first I thought we
Oracle 12c (version 18.104.22.168) offers the option of using in-memory processing to speed things along. Called the In-memory option it’s installed when you install the 22.214.171.124 software. Using it can make a considerable difference in processing speed, provided you have sufficient resources (RAM) available. Let’s revisit an older example, on Bloom filters, and see if Oracle processes things any faster in-memory.
Looking again at the Bloom filter example using (more...)
Inspired by the recent furore around Shellshock I decided that it was time to try an alternative to bash. By the very grown up process of shutting my eyes and poking my finger at the results of a search for “shell” I ended up at fish shell, described by the project as “a command line shell for the 90s”. I’m presuming that this means the 1990s and is not a minimum age requirement.
An interesting issue presented itself just recently with a logical standby database I manage. The database is used for generating reports and the client wanted to skip all DML activity for a given schema as it wasn’t necessary for reporting purposes. I had done this in version 10.2.0.x; it was a simple procedure on a low-traffic database:
alter database stop logical standby apply;
alter database start logical standby apply;
I have installed SQL Developer (with JDK) on Windows many times before, and almost all the time I receive the system error bellow when trying to execute it by the first time. Due to this, I decided to publish the…