Documentation bug on Dynamic Large Object Support - Direct API section
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Object Storage (swift) |
Fix Released
|
Undecided
|
Unassigned |
Bug Description
I've struggled to understand the way I should have named the segments when using dynamic large object upload.
From http://
# First, upload the segments
curl -X PUT -H 'X-Auth-Token: <token>' \
http://<storage_
curl -X PUT -H 'X-Auth-Token: <token>' \
http://<storage_
curl -X PUT -H 'X-Auth-Token: <token>' \
http://<storage_
# Next, create the manifest file
curl -X PUT -H 'X-Auth-Token: <token>' \
-H 'X-Object-Manifest: container/
http://<storage_
# And now we can download the segments as a single object
curl -H 'X-Auth-Token: <token>' \
http://<storage_
Now, if I use this example to upload 10 segments, when I'll download the single object the data will be corrupted, as the segments are concatenated following cardinal order not numerical order. That's why it actually works by adding 000001, 00002 (zeros in front of the number), but it is really confusing and I think it should be clearly explained in the documentation.
I understand in the same page is possible to find the following information, but I'd like to understand it from the example and without checking how's the segmentation/
"swift uses a strict convention for its segmented object support. In the above example it will upload all the segments into a second container named test_container_
I'm available to change the documentation if this make sense.
Many thanks,
Fausto
Changed in swift: | |
assignee: | nobody → janonymous (janonymous) |
Changed in swift: | |
assignee: | janonymous (janonymous) → nobody |
Changed in swift: | |
milestone: | none → 2.4.0 |
status: | Fix Committed → Fix Released |
Suggested fix is to change the example from 1, 2, 3 to 000001, 000002, 000003?
Seems reasonable.