[SERVER-43070] ZSTD fails to decompress messages for versions < v0.8.0 Created: 28/Aug/19 Updated: 27/Oct/23 Resolved: 29/Aug/19 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | Clyde Bazile III (Inactive) | Assignee: | DO NOT USE - Backlog - Dev Tools |
| Resolution: | Works as Designed | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||
| Assigned Teams: |
Developer Tools
|
||||||||
| Operating System: | ALL | ||||||||
| Participants: | |||||||||
| Description |
|
Zstd versions less than v0.8.0 fails to decompress messages and gives the server-side error:
on Ubuntu 16.04. This was discovered with libmongoc, since it was using whichever version of zstd was installed on the system, and Ubuntu 16.04 comes with version 0.5.1 by default. |
| Comments |
| Comment by Kelsey Schubert [ 29/Aug/19 ] |
|
Thanks kevin.albertson, I'm going to close this as WaD since there isn't any Server work to do. Do you know if any other drivers might have similar issues? |
| Comment by Kevin Albertson [ 29/Aug/19 ] |
|
We'll add the version check for libmongoc. But wanted to run this by in case it was unexpected. |
| Comment by Andrew Morrow (Inactive) [ 29/Aug/19 ] |
|
I think you could add a version check to the C driver's CMake find_package or whatever for zstd, and just not compile it in unless you are 0.8.0 or better. Then I think it wouldn't offer it in the negotiation phase at all. |
| Comment by Kelsey Schubert [ 29/Aug/19 ] |
|
Are there any changes that we can make to libmongoc that would make it more user friendly? |
| Comment by Kelsey Schubert [ 29/Aug/19 ] |
|
So it sounds like this is Working as Designed, and we need to clarify this requirement in our documentation for our drivers. Is that correct? |
| Comment by Brian Lane [ 29/Aug/19 ] |
|
Version 0.8.0 is the version where the compression format was finalized which would explain this. |
| Comment by Kelsey Schubert [ 28/Aug/19 ] |
|
FYI brian.lane |