Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Index table medata is not cached during index update #362

Closed
robertpang opened this issue Jul 6, 2018 · 1 comment
Closed

Index table medata is not cached during index update #362

robertpang opened this issue Jul 6, 2018 · 1 comment
Assignees
Labels
kind/enhancement This is an enhancement of an existing feature

Comments

@robertpang
Copy link
Contributor

When a tablet server updates an index of a table being updated, the index's metadata is not cached where it should.

@robertpang robertpang added the kind/enhancement This is an enhancement of an existing feature label Jul 6, 2018
@robertpang robertpang self-assigned this Jul 6, 2018
yugabyte-ci pushed a commit that referenced this issue Jul 6, 2018
Summary: Cache index table metadata in tserver when updating an index for performance. The cache is invalidated when the indexed table is altered.

Test Plan:
```
java -jar ~/code/yugabyte/java/yb-loadtester/target/yb-sample-apps.jar -workload CassandraSecondaryIndex -num_threads_read 0 -num_threads_write 4 -nodes 127.0.0.1:9042
```
Without cache:

```
2018-07-05 17:07:41,973 [INFO|com.yugabyte.sample.common.metrics.MetricsTracker|MetricsTracker] Read: 0.00 ops/sec (0.00 ms/op), 0 total ops  |  Write: 215.81 ops/sec (18.53 ms/op), 3517 total ops  |  Uptime: 25015 ms | maxWrittenKey: 3514 | maxGeneratedKey: 3520 |
2018-07-05 17:07:46,976 [INFO|com.yugabyte.sample.common.metrics.MetricsTracker|MetricsTracker] Read: 0.00 ops/sec (0.00 ms/op), 0 total ops  |  Write: 217.65 ops/sec (18.37 ms/op), 4606 total ops  |  Uptime: 30018 ms | maxWrittenKey: 4605 | maxGeneratedKey: 4609 |
2018-07-05 17:07:51,979 [INFO|com.yugabyte.sample.common.metrics.MetricsTracker|MetricsTracker] Read: 0.00 ops/sec (0.00 ms/op), 0 total ops  |  Write: 212.29 ops/sec (18.82 ms/op), 5668 total ops  |  Uptime: 35021 ms | maxWrittenKey: 5665 | maxGeneratedKey: 5671 |
```

With cache:

```
2018-07-05 17:12:24,773 [INFO|com.yugabyte.sample.common.metrics.MetricsTracker|MetricsTracker] Read: 0.00 ops/sec (0.00 ms/op), 0 total ops  |  Write: 229.69 ops/sec (17.42 ms/op), 2892 total ops  |  Uptime: 20015 ms | maxWrittenKey: 2891 | maxGeneratedKey: 2895 |
2018-07-05 17:12:29,776 [INFO|com.yugabyte.sample.common.metrics.MetricsTracker|MetricsTracker] Read: 0.00 ops/sec (0.00 ms/op), 0 total ops  |  Write: 231.24 ops/sec (17.31 ms/op), 4049 total ops  |  Uptime: 25018 ms | maxWrittenKey: 4048 | maxGeneratedKey: 4052 |
2018-07-05 17:12:34,781 [INFO|com.yugabyte.sample.common.metrics.MetricsTracker|MetricsTracker] Read: 0.00 ops/sec (0.00 ms/op), 0 total ops  |  Write: 235.37 ops/sec (16.97 ms/op), 5227 total ops  |  Uptime: 30023 ms | maxWrittenKey: 5225 | maxGeneratedKey: 5230 |
```

Also verify that no further RPC is executed at the master leader when the workload is run.

Reviewers: pritam.damania, amitanand, mihnea

Reviewed By: mihnea

Subscribers: ybase, yql

Differential Revision: https://phabricator.dev.yugabyte.com/D5112
@robertpang
Copy link
Contributor Author

Fixed in commit 2ebadb8.

jasonyb pushed a commit that referenced this issue Jun 11, 2024
1) Added the Group by clause to make sure that bucket change doesn't have any
impact on aggregates of queries.
2) Updated column names where required.
3) Updated pgbench parameters to reduce the time taken by test case, around
70-80% decrease in time taken by test case.
jasonyb pushed a commit that referenced this issue Jul 10, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
kind/enhancement This is an enhancement of an existing feature
Projects
None yet
Development

No branches or pull requests

1 participant