-
Notifications
You must be signed in to change notification settings - Fork 261
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 1.8 #3132
Comments
Signed-off-by: Nigel Jones <nigel.l.jones+git@gmail.com>
Signed-off-by: Nigel Jones <nigel.l.jones+git@gmail.com>
Signed-off-by: Nigel Jones <nigel.l.jones+git@gmail.com>
Signed-off-by: Nigel Jones <nigel.l.jones+git@gmail.com>
Signed-off-by: Nigel Jones <nigel.l.jones+git@gmail.com>
Signed-off-by: Nigel Jones <nigel.l.jones+git@gmail.com>
Signed-off-by: Nigel Jones <nigel.l.jones+git@gmail.com>
#3132 updates to readme for new versioning, 1.8 being current
CTS (graph and inmem) complete with no errors Profile results:
|
Testing the 'cohort' notebooks indicates some incomplete cohort registration. Not all servers are seeing all others. Specifically cocoMDS4 is not reported as a remote member by any other members of the cohort. This is because cocoMDS4 does not have a metadata repository - and will not receive any federated queries. This isn't a regression - but could be confusing and may need doc clarification. It may also be useful to be able to see all servers that have registered even if not a member of the cohort - for example so we can monitor those servers, and query configured services (such as OMASs) from a bootstrap entry point cc: @mandy-chessell |
#2741 will look at improving the cohort registration process for members without a repo |
Potential issue in vdc helm chart
This error does not occur in master The cause of this is due to the strings vs integers when parsing the template In the dev build (ie master) we use '2.0-SNAPSHOT' which is interpreted as a string, whilst a simple '1.8' or '2.0' is not. This then affects the template parsing when creating the config map. |
Signed-off-by: Nigel Jones <nigel.l.jones+git@gmail.com>
Signed-off-by: Nigel Jones <nigel.l.jones+git@gmail.com>
#3132 Fix use of version in vdc template
RELEASE 1.8 #3132 Fix use of version in vdc template
With testing completed I kicked off the build pipeline yesterday. We're now on our third sync attempt (artifactory->bintray->jcenter->maven central) but maven central is seeing issues this week as per https://status.maven.org & is currently under maintainance. |
The following artifacts cannot be released:
I have contacted JFrog support for rex-view (cc: @grahamwallis ) though since it's unlikely any egeria consumers would be dependent on this artifact - and it's only a packaging pom the impact is likely minor. |
Thanks for the heads-up @planetf1 re: rex-view. It sounds strange - it's just an implementation of a view service and should look very similar to glossary-author. I can't spot the difference yet. Quite soon tex-view is one the way, so it will be interesting to see if JCenter complains about that VS too. |
After looking at this more closely, I think the issue actually is that both rex-view & data-engine-proxy-services-server never got as far as bintray,. The former is not in JCenter since it's a new module, the latter is ... but this is not the problem in itself. Rather there is no built 1.8 version to stage. I've gone through the release logs - and I can see rex-view being built, and indeed loaded into artifactory. The 'promotion' of repos is at a repo level so either all or none work. The sync to bintray is then a single jfrog cli command which succeeded - yet it seems as if this is where rex-view was 'lost'. No errors reported. This is a stage BEFORE where we've seen issues to-date which is in the bintray->JCente->maven central pipeline. With rex-view it never even made it that far One of the challenges with this process is we only do it infrequently - in future perhaps publishing occasional snapshots through the same process would help flush out issues quicker. Also I don't have all the necessary credentials to clean things up. I will raise an issue at the LF to try and review the process and look at closing some of these gaps we have in place -- for example the jobs perhaps should do verification beyond checking the api succeeded by explicit checks |
For now I don't think the omission of these two artifacts will impact users. However I'll try and manually upload to bintray->JCenter. I could rerun the full pipeline but will likely hit permission problems, and if not could spin another few days on the sync as the jobs aren't setup for small incremental updates. Given the nature of the omission this does not seem warranted. |
Summarised issues in #3194 - for future improvements in the reliability of this process data-engine-proxy-services-server uploaded manually (already in JCenter) and synced to MC |
rex-view now synced too. |
Proposal to create Release 1.8 branch 2 June 2020, and then switch master to 2.0-SNAPSHOT
The 2.0 switch is due in part to the work (#1403 ) to switch to using https/ssl for all REST APis
The text was updated successfully, but these errors were encountered: