There are some new classes here that represent a new version of the original simple program. I have taken the original SimpleBucketFiller and refactored it so that there is now a class to represent a test configuration, test results, a factory, and a configuration consumer. The main, runnable class is SBFConfigurationConsumer. When it starts up it gets a list of test configurations, SBFConfiguration objects, from the SBFConfigurationFactory. Then it iterates over them, performs the test according to the specified options, and records the results in SBFTestResults. There is also now support for durability options.
Sample output
Hostname: 10.111.110.101
Persist To: NONE
Replicate To: NONE
SDK version is: couchbase-java-client/2.2.6 (git: 2.2.6, core: 1.2.7)
Successfully inserted: 10 documents.
# of documents that already existed: 0
Number of durability exceptions: 0
Number of timeout exceptions: 0
Number of other exceptions: 0
Total run time was: 697 millseconds
Performance was about: 14 docs per second
Total successful doc size was: 9179977 bytes
Generated doc size range was between 834757 and 985248 bytes
Note: As a single-threaded, synchronous, application, one single instance of this program will not come close to reaching the maximum ops/sec capacity of most clusters.
Utility for inserting documents into a bucket with random expiration times and random sizes.
Sample output
Done. Successfully inserted: 1000 documents.
# of documents that already existed: 0
Number of other exceptions: 0
Total run time was: 1312 millseconds
Performance was about: 762 docs per second
Total successful doc size was: 85125 bytes
Generated doc size range was between 21 and 90 bytes
"expiry" property is now added into the document. So if you use this utility to push code into Couchbase 4.x, you can use N1QL to then query for a histogram of the document expiration times. For example, if you want to see document expirations broken down into 100 second buckets:
select trunc(expiry / 100) AS expBucket, count(*)
AS bucketSize from `BUCKETNAME`
group by trunc(expiry/100)
order by trunc(expiry/100);
The results will look something like this
"results": [
{
"bucketSize": 25,
"expBucket": 33
},
{
"bucketSize": 11,
"expBucket": 34
},
{
"bucketSize": 26,
"expBucket": 35
}
],
Utility to get each document in the bucket, where the keys can be programmatically determined. Can raise the resident ratio of a bucket.
Sample output
Done. Successfully gotten: 166 documents out of 1000 tried.
Null document count: 834
Number of other exceptions: 0
Total run time was: 626 millseconds
Performance was about: 1597 docs per second