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

Release 0.17.0 #1355

Closed
busbey opened this issue Sep 21, 2019 · 30 comments
Closed

Release 0.17.0 #1355

busbey opened this issue Sep 21, 2019 · 30 comments

Comments

@busbey
Copy link
Collaborator

busbey commented Sep 21, 2019

changes from 0.16.0 (#1265) that need testing:

core changes

added bindings

updated bindings

removed bindings

@busbey
Copy link
Collaborator Author

busbey commented Sep 21, 2019

Release candidate 1 is up:
https://github.com/brianfrankcooper/YCSB/releases/tag/0.17.0-RC1

As a reminder, testing a given binding should at a minimum run through a load and a run phase. Ideally, one should go through the recommended sequence of workload steps from the "Running the workloads" section of the Core Workloads documentation.

@busbey
Copy link
Collaborator Author

busbey commented Sep 23, 2019

I confirmed that the change to limit the thread count works. and I checked an HDP 2.6.5 HBase cluster with the hbase098, hbase10, hbase12, hbase14, and aysnchbase bindings.

@busbey
Copy link
Collaborator Author

busbey commented Sep 23, 2019

I confirmed that the hbase20 binding works with a CDH6 HBase cluster.

@busbey
Copy link
Collaborator Author

busbey commented Sep 23, 2019

oh and I confirmed that the fieldnameprefix stuff works.

@busbey
Copy link
Collaborator Author

busbey commented Sep 23, 2019

Things that it'd be nice to test prior to marking the release as done, so that they don't need to be marked as experimental:

@voellm
Copy link
Contributor

voellm commented Sep 23, 2019

We regularly build and test both the Azure CosmoDB and Cassandra + SSL changes and they work as expected. What kind of update / test output / etc would you like?

@busbey
Copy link
Collaborator Author

busbey commented Sep 23, 2019

We regularly build and test both the Azure CosmoDB and Cassandra + SSL changes and they work as expected. What kind of update / test output / etc would you like?

Doing those same checks for the 0.17.0 RC tag specifically would be great. Generally we take folks who test at their word rather than requiring a copy of any results.

@KowsarAtz
Copy link
Contributor

KowsarAtz commented Sep 24, 2019 via email

@srmadscience
Copy link
Contributor

"@srmadscience mind checking that the voltdb stuff works in the release candidate?".

Done for VoltDB v9.1.1.

Question: Any idea when this is 'officially' released?

@busbey
Copy link
Collaborator Author

busbey commented Sep 24, 2019

Question: Any idea when this is 'officially' released?

I'll wait a couple of days to make sure no one else is testing any of the outstanding stuff. Then I'll post a release and link it here. This issue will be closed once download links, etc get updated.

@patrickstuedi
Copy link
Contributor

Done. Ran the basic tests for Crail against the 0.17.0 RC and things work fine.

@leschekhomann
Copy link
Contributor

@busbey: At the moment I am busy, but I try to test the binding during the next week.

@knonomura
Copy link
Contributor

@busbey
Done for GridDB.
I can build and run successfully 0.17.0 RC tag.

@voellm
Copy link
Contributor

voellm commented Sep 25, 2019

I have built and run the Azure Cosmos [SQL] workloads and well as testing connections to Azure Cosmos Cassandra using the useSSL flag.

@allanbank
Copy link
Collaborator

Ran my test script for MongoDB versions:

  • 4.2.0
  • 4.0.12
  • 3.6.14
  • 3.4.23
  • 3.2.22
  • 3.0.15
  • 2.6.12

No issues.

@rizlik
Copy link

rizlik commented Sep 26, 2019

I can test this in the beginning of the next week.
btw as far as I'm understanding just one of the issue is addressed. As the main one remains unsolved isn't better to re-open the issue (#1153 )?
Also does the, now fixed, data integrity triggers an error without fixing the wrong update logic?

@busbey
Copy link
Collaborator Author

busbey commented Sep 26, 2019

I can test this in the beginning of the next week.
btw as far as I'm understanding just one of the issue is addressed. As the main one remains unsolved isn't better to re-open the issue (#1153 )?
Also does the, now fixed, data integrity triggers an error without fixing the wrong update logic?

I'd love someone to test and find out the answers to these questions. Either reopening or a new issue would be fine, if there's still problems.

@rizlik
Copy link

rizlik commented Sep 28, 2019

re memcahed:

e6bd739 looks ok but it doesn't fix the issues. In particular on an update operation all the fields but the updated one are destroyed and dataintegrity doesn't catch this. I think it would be better to open two new separated issues. I attach a raw simple script that show the wrong behavior (sorry for the wrong extension but github doesn't allow me to attach .sh).

If it looks sound to you I can open the issues and add some minor details.
memcached_test.sh.txt

@busbey
Copy link
Collaborator Author

busbey commented Sep 28, 2019

Yep, your analysis looks correct. Open the two issues and link them here please? I'll list them as known issues for the release.

@rizlik
Copy link

rizlik commented Sep 29, 2019

Yep, your analysis looks correct. Open the two issues and link them here please? I'll list them as known issues for the release.

Here they are: #1358 and #1359

@seybi87
Copy link
Contributor

seybi87 commented Oct 1, 2019

run several load/run phase tests for with the following auto-failover JDBC connection string with the
PostgresSQL JDBC Driver against CockroachDB 2.0.3
jdbc:postgresql://IP1:PORT1, IP2:PORT2,IP3:PORT3/ycsb?sslmode=disable;jdbc:postgresql://IP1:PORT1, IP2:PORT2,IP3:PORT3/ycsb?sslmode=disable

Execution works without any problems.

Applied the load phase in combination with a node failure to test the auto-failover explicitly:
One YCSB thread seems to be entering a non-recoverable error state and is not terminating but keeps on reporting 0 ops/s.
Yet, I did not have time to investigate further if this behavior is caused by CockroachDB or if there is an issues in the JDBC binding (it worked with the same delimiter changes applied to the former 0.12.0 YCSB version)

Anyways, the load/run phase execution works without issues if the newly enabled auto-failover capabilites are not used.

@isuntsov-gridgain
Copy link
Contributor

isuntsov-gridgain commented Oct 1, 2019 via email

@busbey
Copy link
Collaborator Author

busbey commented Oct 2, 2019

I'm going to add #1356 as a known issue and wrap up this release in the next day or so.

@busbey
Copy link
Collaborator Author

busbey commented Oct 6, 2019

the release is up

I'm going to work through the steps for publishing to maven central (for core and binding-parent) before closing this issue out.

let me know if the release notes look like they're missing anything.

@leschekhomann
Copy link
Contributor

Hi @busbey,
I tested the Postgres NoSQL binding against the official docker containers of PostgreSQL 9.5 and 9.6 from https://hub.docker.com/_/postgres. I run the sequence with the default parameters and workloads as described in "Running the workloads" without changing the number of threads or targets. At the moment the execution of the sequence only works when a property file is provided by the user since the URL, user name and user password are empty strings by default. Furthermore, an exception is thrown when the target database does not exist.

@busbey
Copy link
Collaborator Author

busbey commented Oct 7, 2019

Thanks for the update! I'll move postgresnosql to the tested set in the release notes.

@HoboChen
Copy link
Contributor

HoboChen commented Oct 8, 2019

As I left alibaba cloud, @RaymondChina will take over the work you mentioned; thanks for your time! @busbey

@RaymondChina
Copy link

for Tablestore, the test pass. @busbey

what I have done :

  1. clone the repo, switch to 0.17.0-RC1
  2. follow the steps in YCSB/Tablestore/README.md, apply account, create instance, create table, fill configuration, start test cases
  3. the result is ok

@busbey
Copy link
Collaborator Author

busbey commented Oct 9, 2019

Thanks @RaymondChina! I updated the release notes.

Note to self: I still need to upload the convenience tarballs for postgrenosql and tablestore.

@busbey
Copy link
Collaborator Author

busbey commented Oct 10, 2019

okay, I added convenience tarballs for postgrenosql and tablestore. also published the stuff needed to make a binding outside of the project on oss.sonatype.org.

Thanks everyone who helped out with testing!

@busbey busbey closed this as completed Oct 10, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests