kubevirt / containerized-data-importer
59%

DEFAULT BRANCH: main
Repo Added
Files 156
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst

LAST BUILD ON BRANCH main
branch: SELECT
CHANGE BRANCH
x

coverage: 59.3% (+0.002%) from 59.298%
#5026

push

travis-ci

web-flow 
CNV-51521 | Fix Incomplete transfer of change blocks leads to data corruption (#3534)

* CNV-51521 | Fix Incomplete transfer of change blocks leads to data corruption

Issue: When user has a large disk with over 2000 changed blocks the
QueryChangedDiskAreas returns only first 2000 changed blocks. The disk
is not getting all the changes form the VMware Change Block Tracking and
this can cause disk corruption.

Fix: Query the changed areas by every 2000 block chunks, combine the chunks
together and request the next QueryChangedDiskAreas to start from the
end of the last block. Once the QueryChangedDiskAreas will return less
than 2000 blocks we know that it was the last chunk and can continue
with the disk import.

Ref: https://issues.redhat.com/browse/CNV-51521

Signed-off-by: Martin Necas <mnecas@redhat.com>

* Fix unit tests for new changed block query behavior.

* Fix unit tests for new changed block query behavior.

Signed-off-by: Matthew Arnold <marnold@redhat.com>

* Fix lint failure.

Signed-off-by: Matthew Arnold <marnold@redhat.com>

---------

Signed-off-by: Matthew Arnold <marnold@redhat.com>

* Get VDDK warm import tests working.

Change the QueryChangedDiskAreas loop to always quit on zero-length
change lists. The previous condition would spin forever when vcsim
returned 0 for every request.

Signed-off-by: Matthew Arnold <marnold@redhat.com>

* Add a limit for the disk capacity before querying the changed blocks

Issue: When running the QueryChangedDiskAreas with startOffset of the
disk size we would get following erorr.
```
Failed to query changed areas: ServerFaultCode: A specified parameter was not correct: startOffset
```

Fix: Add query to get the VMware disk size to compare the offset and the
disk size.

Signed-off-by: Martin Necas <mnecas@redhat.com>

---------

Signed-off-by: Martin Necas <mnecas@redhat.com>
Signed-off-by: Matthew Arnold <marnold@redhat.com>
Co-authored-by: Matthew Arnold <5075485+mrnold@users.noreply.git... (continued)

42 of 57 new or added lines in 1 file covered. (73.68%)

11 existing lines in 1 file now uncovered.

16735 of 28221 relevant lines covered (59.3%)

0.66 hits per line

Relevant lines Covered
28221 RELEVANT LINES 16735 COVERED LINES
0.66 HITS PER LINE
Source Files on main

Recent builds

Builds Branch Commit Type Ran Committer Via Coverage
#5026 main CNV-51521 | Fix Incomplete transfer of change blocks leads to data corruption (#3534) * CNV-51521 | Fix Incomplete transfer of change blocks leads to data corruption Issue: When user has a large disk with over 2000 changed blocks the QueryChange... push web-flow travis-ci
59.3
#5025 v1.60.4 v1.60.4 release notes Signed-off-by: Alex Kalenyuk <akalenyu@redhat.com> push akalenyu travis-ci
59.19
#5024 main cdi-uploadserver: healthcheck and application use same port (#3536) healthcheck used to be on a dedicated port for "reasons" and by that I mean I was too lazy to properly set up the tls configuration and check authz in upload handlers Signed-off... push web-flow travis-ci
59.3
#5023 CNV-51521 Add a limit for the disk capacity before querying the changed blocks Issue: When running the QueryChangedDiskAreas with startOffset of the disk size we would get following erorr. ``` Failed to query changed areas: ServerFaultCode: A specified par... Pull #3534 mnecas travis-ci
59.34
#5022 CNV-51521 Add a limit for the disk capacity before querying the changed blocks Issue: When running the QueryChangedDiskAreas with startOffset of the disk size we would get following erorr. ``` Failed to query changed areas: ServerFaultCode: A specified par... Pull #3534 mnecas travis-ci
59.31
#5021 CNV-51521 Get VDDK warm import tests working. Change the QueryChangedDiskAreas loop to always quit on zero-length change lists. The previous condition would spin forever when vcsim returned 0 for every request. Signed-off-by: Matthew Arnold <marnold@redha... Pull #3534 mrnold travis-ci
59.31
#5020 uploadserver-health cdi-uploadserver: healthcheck and application use same port healthcheck used to be on a dedicated port for "reasons" and by that I mean I was too lazy to properly set up the tls configuration and check authz in upload handlers Signed-off-by: Mic... Pull #3536 mhenriks travis-ci
59.29
#5019 CNV-51521 Fix unit tests for new changed block query behavior. * Fix unit tests for new changed block query behavior. Signed-off-by: Matthew Arnold <marnold@redhat.com> * Fix lint failure. Signed-off-by: Matthew Arnold <marnold@redhat.com> ----... Pull #3534 web-flow travis-ci
59.3
#5018 CNV-51521 Fix unit tests for new changed block query behavior. * Fix unit tests for new changed block query behavior. Signed-off-by: Matthew Arnold <marnold@redhat.com> * Fix lint failure. Signed-off-by: Matthew Arnold <marnold@redhat.com> ----... Pull #3534 web-flow travis-ci
59.3
#5017 CNV-51521 CNV-51521 | Fix Incomplete transfer of change blocks leads to data corruption Issue: When user has a large disk with over 2000 changed blocks the QueryChangedDiskAreas returns only first 2000 changed blocks. The disk is not getting all the change... Pull #3534 mnecas travis-ci
59.3
See All Builds (5028)