[
https://issues.apache.org/jira/browse/SLING-7234?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16284348#comment-16284348
]
Karl Pauls commented on SLING-7234:
-----------------------------------
[~olli], that looks like a bug. However, I'm not sure that this is related to
this bug because that is about a cycle while in your case it is about a none
existing config that doesn't get handled correctly. I think for now (i.e.,
until the bug is fixed) you have two options namely, a) create a ScriptCache
configuration similar to the one in the sling starter [0] (it needs to have at
least one value for the additionalExtensions) or b) update to the latest R7 SCR
SNAPSHOT (2.1.0-SNAPSHOT) which might fix this (but will not be released for at
least another month or two).
[~radu.cotescu], if you call a property method on a config object that returns
an array it will return null if there is no config (regardless of whether you
set a default of {} or not). Thats one spec oddity of the old DS spec (it will
be changed for the new DS). As a consequence, it looks like [1] is blowing up
because there is no config (or it doesn't have a value for addtionalExtensions)
and hence, additionalExtensions is set to null in [2].
[0]
https://github.com/apache/sling-org-apache-sling-starter/blob/master/src/main/provisioning/scripting.txt#L54
[1]
https://github.com/apache/sling-org-apache-sling-scripting-core/blob/master/src/main/java/org/apache/sling/scripting/core/impl/ScriptCacheImpl.java#L251
[2]
https://github.com/apache/sling-org-apache-sling-scripting-core/blob/master/src/main/java/org/apache/sling/scripting/core/impl/ScriptCacheImpl.java#L219
> ServiceFactory.getService() resulted in a cycle in sling karaf
> ---------------------------------------------------------------
>
> Key: SLING-7234
> URL: https://issues.apache.org/jira/browse/SLING-7234
> Project: Sling
> Issue Type: Bug
> Components: Karaf, Scripting
> Affects Versions: Scripting Core 2.0.50
> Environment: Debian linux, jdk 1.8.0_151-b12
> Reporter: Ivo Leitão
> Assignee: Oliver Lietz
> Attachments: clean.log.1, grep-errors.log, sample.log.1
>
>
> Hello,
> Once in a while on the first startup of Sling on Karaf I'm seeing the
> exception ServiceFactory.getService() resulted in a cycle. This does not
> happen every time only on some startups.
> After the first startup on the next restart this exception does not appear
> anymore.
> I've picked a sample of the exception. Unfortunately this seems like a
> concurrency issue and therefore much harder to find a pattern but it happens
> mostly with sling scripting. I've seen other exceptions also with sling
> scripting but this one is the most common
> I've uploaded two log files:
> clean.log.1 with a clean startup and
> sample.log.1 with the aforementioned exception
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
