cloudstack-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <>
Subject [jira] [Commented] (CLOUDSTACK-8703) Cloudstack tries to re-download templates on S3 while they are already their.
Date Mon, 03 Aug 2015 16:30:04 GMT


ASF GitHub Bot commented on CLOUDSTACK-8703:

Github user DaanHoogland commented on a diff in the pull request:
    --- Diff: utils/src/com/cloud/utils/ ---
    @@ -352,10 +352,15 @@ public static File getFile(final ClientOptions clientOptions, final
String bucke
             ListObjectsRequest listObjectsRequest = new ListObjectsRequest().withBucketName(bucketName).withPrefix(directory
             ObjectListing ol = client.listObjects(listObjectsRequest);
    --- End diff --
    we don't catch any of the amazon runtime exceptions. I think we should wrap them in a
CloudstackRuntimeException so they can be caught be the servlet before propagating into the
container. This is not in your PR @borisroman so I won't -1 on it but if you are on it, why
not improve ;)

> Cloudstack tries to re-download templates on S3 while they are already their.
> -----------------------------------------------------------------------------
>                 Key: CLOUDSTACK-8703
>                 URL:
>             Project: CloudStack
>          Issue Type: Bug
>      Security Level: Public(Anyone can view this level - this is the default.) 
>          Components: Secondary Storage
>    Affects Versions: 4.5.2
>            Reporter: Boris Schrijver
>            Assignee: Boris Schrijver
>             Fix For: 4.6.0, 4.5.2
> When the management server restarts, reason doesn't matter, it will try to list all of
the available templates on the secondary storage. To redownload them when they are missing.
On S3, S3Utils listDirectory() will not return these because of a bug introduced in the following
commit: debcbcc1454789936407583fe0aba8e158b16cad. So the management server will redownload
them all. The listDirectory() method in will only return objects when the anwser
it receives from the S3 system is truncated. So listing smaller then the pagination limit
will always return 0 obejcts.

This message was sent by Atlassian JIRA

View raw message