Opened 13 years ago
Closed 13 years ago
#1367 closed task (fixed)
Provide a clear upgrade path to 2.0
Reported by: | strk | Owned by: | pramsey |
---|---|---|---|
Priority: | medium | Milestone: | PostGIS 2.0.0 |
Component: | postgis | Version: | master |
Keywords: | Cc: |
Description
We need a way to upgrade an existing spatial database to 2.0 w/out having to ignore errors here and there.
Ideally a script could take care of it all, but first thing is defining the process steps.
Attachments (1)
Change History (8)
comment:1 by , 13 years ago
comment:2 by , 13 years ago
By the end of the process above I end up with some tables using geometry column constraints and some tables using typmod. We may need a step 4. to perform the cleanup/conversion or embed that part directly into step 3. (postgis_restore phase). Splitting would help debugging and be probably safer.
by , 13 years ago
Attachment: | import_report16dec11.txt added |
---|
comment:3 by , 13 years ago
I have now tried this once on a non-trivial database. It appears to succeed.
source database ⇒ POSTGIS="1.5.3" GEOS="3.2.2-CAPI-1.6.2" PROJ="Rel. 4.7.1, 23 September 2009" LIBXML="2.7.7" USE_STATS
destination database, trunk rev 8447 ⇒ POSTGIS="2.0.0SVN" GEOS="3.4.0dev-CAPI-1.8.0" PROJ="Rel. 4.7.1, 23 September 2009" LIBXML="2.7.6" USE_STATS
comment:6 by , 13 years ago
With r8468 the postgis_restore.pl script also takes care of spatial_ref_sys by restoring all entries in dump with SRIDs not found in pre-existing spatial_ref_sys table, thus allowing you to decide which version of the core srid you want (you probably want the new ones).
This means that step 2 in the first comment (spatially-enable new database) could now include sourcing spatial_ref_sys.sql w/out worrying about getting a zilion "duplicate primary key" errors at restore time.
I guess all is left here is documenting the procedure and calling for test.
comment:7 by , 13 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
r8469 documents the procedure in the existing HARD UPGRADE section of the PostGIS manual. Any left over I would call a bug, thus worth a separate ticket. Thanks for testing!
My current idea is:
I've been working on making this model work. Can you see conceptual issues with it ?