one icon indicating copy to clipboard operation
one copied to clipboard

Successfull but incomplete image import from HTTP

Open vholer opened this issue 4 years ago • 0 comments

Description

It can happen (don't know how to reproduce) that image being imported into datastore from remote HTTP server is not downloaded completely (or broken after download?, maybe via fallocate), but still imported successfully at the end. Such image is obviously broken. Example of imported and original image:

-rw-r--r-- 1 root root 652214272 Apr 14 11:49 b3c5bde0eeacbaae193c4a42a790ba0a
-rw-r--r-- 1 root root 652607488 Nov  9 11:11 fedora31.qcow2

(difference of ~384 KiB)

Tue Apr 13 18:34:28 2021 [Z0][ReM][D]: Req:5584 UID:0 IP:127.0.0.1 one.image.allocate invoked , "NAME="Fedora 31"
TYPE="OS"
PATH="http://services/images/fedora31.qcow2"
FORMAT="qcow2"
", 1, false
Tue Apr 13 18:34:28 2021 [Z0][ReM][D]: Req:752 UID:0 IP:127.0.0.1 one.datastorepool.info invoked 
Tue Apr 13 18:34:28 2021 [Z0][Lis][I]: Stopping 
Tue Apr 13 18:34:28 2021 [Z0][ImM][I]: Copying http://services/images/fedora31.qcow2 to repository for image 1
Tue Apr 13 18:34:28 2021 [Z0][ReM][D]: Req:5584 UID:0 one.image.allocate result SUCCESS, 1

When I retry the import now, it imports identical image.

To Reproduce Unknown, making the issue for the record.

Details

  • Affected Component: Storage
  • Hypervisor: n/a
  • Version: 6.0.0

Progress Status

  • [ ] Branch created
  • [ ] Code committed to development branch
  • [ ] Testing - QA
  • [ ] Documentation
  • [ ] Release notes - resolved issues, compatibility, known issues
  • [ ] Code committed to upstream release/hotfix branches
  • [ ] Documentation committed to upstream release/hotfix branches

vholer avatar Apr 14 '21 12:04 vholer