[
http://mifosforge.jira.com/browse/MIFOSADMIN-163?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Adam Monsen updated MIFOSADMIN-163:
-----------------------------------
Description:
The current hudson jobs should follow best practices in order to ensure build
integrity/reproducibility.
# Wipe out workspace or Clean after checkout
# Use private maven repository
See this build for example in which the above are not configured:
https://ci.mifos.org/hudson/job/head-master/configure
This will ensure artifacts that are installed from one builds does not leak
into another build.
was:
The current hudson jobs should follow best practices in order to ensure build
integrity/reproducibility.
1) Wipe out workspace or Clean after checkout
2) Use private maven repository
See this build for example in which the above are not configured:
https://ci.mifos.org/hudson/job/head-master/configure
This will ensure artifacts that are installed from one builds does not leak
into another build.
Implied subtask: publish snapshot artifacts such as mifos-api (or give them
release versions, and publish those)
Priority: Critical (was: Major)
Recently we've learned that this has a huge impact on developer productivity.
Udai has been working on speeding up the build by creating a slave ci server.
Allowing head-master to run on the slave required a change in the build
configuration: not archiving artifacts. Archiving basically keeps backups of
build products (like the .war and .zip). But when head-master is built on a
slave, the artifacts must be copied back to the master ci server, and this
takes a very long time. What Udai did to overcome this was create a
head-master-archive job. head-master-archive *re-runs* the entire build,
skipping tests, and archives the artifacts. head-master-archive is tied to the
master ci server, so we know it will be fast.
However, there are problems with this approach:
* we're promoting untested build products to test servers
* if someone commits during the head-master build, the head-master-archive will
build *different* code
My proposed solution (captured in MIFOSADMIN-291) is to, at every build step,
*deploy* artifacts into our nexus maven repository. Not only will this make our
build(s) more robust (we only build and test a particular binary once), this
will also speed up every build that relies on the same artifacts, *including
builds on every developer workstation*!
> Robust CI builds
> ----------------
>
> Key: MIFOSADMIN-163
> URL: http://mifosforge.jira.com/browse/MIFOSADMIN-163
> Project: mifos administration
> Issue Type: Improvement
> Reporter: Sumit Shah
> Assignee: mifosdeveloperqueue
> Priority: Critical
> Fix For: Cloud Release A, Cloud Release A - Unscheduled
>
>
> The current hudson jobs should follow best practices in order to ensure build
> integrity/reproducibility.
> # Wipe out workspace or Clean after checkout
> # Use private maven repository
> See this build for example in which the above are not configured:
> https://ci.mifos.org/hudson/job/head-master/configure
> This will ensure artifacts that are installed from one builds does not leak
> into another build.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://mifosforge.jira.com/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
------------------------------------------------------------------------------
Special Offer-- Download ArcSight Logger for FREE (a $49 USD value)!
Finally, a world-class log management solution at an even better price-free!
Download using promo code Free_Logger_4_Dev2Dev. Offer expires
February 28th, so secure your free ArcSight Logger TODAY!
http://p.sf.net/sfu/arcsight-sfd2d
_______________________________________________
Mifos-issues mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mifos-issues
