Disk usage

[Matthias Sohn <matthias.sohn@xxxxxxxxx>]

On Mon, Jun 10, 2013 at 8:28 AM, Mickael Istria <mistria@xxxxxxxxxx> wrote:

On 06/10/2013 08:25 AM, Matthias Sohn wrote:

Since sandbox hudson emits Bad Gateway errors in random situations I am going to restart it hoping this will fix these problems.

Hi Matthias,

You should first have a look at this issue: https://bugs.eclipse.org/bugs/show_bug.cgi?id=410271 .
If you need some free space, feel free to remove all builds from SWTBot-Gerrit and SWTBot-Sonar jobs (except the last successful one). Not sure it will be enough though.

I restarted sandbox Hudson but still the egit.gerrit job faile with a "Bad Gateway" error

(https://hudson.eclipse.org/sandbox/job/egit.gerrit/4622/console):

Caused by: org.apache.maven.wagon.TransferFailedException: Failed to transfer file: http://repo.maven.apache.org/maven2/org/codehaus/gmaven/feature/gmaven-feature-support/1.4/gmaven-feature-support-1.4.jar. Return code is: 502, ReasonPhrase:Bad Gateway.
	at org.apache.maven.wagon.shared.http4.AbstractHttpClientWagon.fillInputData(AbstractHttpClientWagon.java:852)
	at org.apache.maven.wagon.StreamWagon.getInputStream(StreamWagon.java:116)
	at org.apache.maven.wagon.StreamWagon.getIfNewer(StreamWagon.java:88)
	at org.apache.maven.wagon.StreamWagon.get(StreamWagon.java:61)
	at org.sonatype.aether.connector.wagon.WagonRepositoryConnector$GetTask.run(WagonRepositoryConnector.java:601)
	... 4 more

Then I configured the SWTBot-Gerrit and SWTBot-Sonar jobs to only keep the artifacts

of the last succeeded build which freed some diskspace. It seems this helped to enable

the egit.gerrit job to run successfully again.

Looking at current disk usage at https://hudson.eclipse.org/sandbox/plugin/disk-usage/

there are a couple of jobs consuming quite a lot of disk space for their workspaces,

maybe some of them could free some space. I'd recommend they all configure their

jobs to only retain the artifacts of the last successful build.

Probably a large fraction of this space is consumed by job-private maven repositories.

Not sure if there is a way to get rid of the need to use private maven repositories, 

AFAIK this is needed since sharing local maven repositories between multiple jobs

doesn't work reliably.

Overall it looks like sandbox Hudson needs more HDD resources.

 Disk usage

Builds:4GB, Workspace:33GB

Project name	Builds	Workspace  ↑
platform-sonar	68MB	7GB
target-management	144MB	2GB
linuxtools	1GB	2GB
linuxtools-sonar	88MB	1GB
mylyn-context-mft-gerrit	878KB	1GB
mylyn-context-gerrit	2MB	1GB
mylyn-reviews-gerrit	16MB	1GB
mylyn-tasks-gerrit	152MB	1GB
xtext.gerrit	703MB	1GB
mylyn-docs-gerrit	14MB	964MB
epp-mpc-gerrit	11MB	963MB
mylyn-commons-gerrit	72MB	805MB
e4-tools	60MB	744MB
GMF-Tooling-Sonar	540MB	725MB
mylyn-builds-gerrit	1MB	718MB
emf-compare.gerrit	3MB	710MB
recommenders.gerrit-e3x	26MB	700MB
recommenders.gerrit	29MB	678MB
mylyn-versions-gerrit	521KB	667MB
tcf.gerrit	26MB	583MB
mylyn-reviews-r4e-gerrit	22MB	570MB
koneki-ldt-sonar	49MB	556MB
vjet-juno-sonar	239MB	552MB
emfstore-sonar	56MB	542MB
eclipse.scout.sdk	29MB	507MB

 

--

Matthias