Index creation issue on blocking parsing of some statefiles

[mvisonneau]

On a specific statefile that my bucket contains, I can see the following errors being thrown by postgres:

STATEMENT:  INSERT INTO "attributes" ("resource_id","key","value") VALUES ($1,$2,$3) RETURNING "attributes"."id"
ERROR:  index row size 5048 exceeds maximum 2712 for index "idx_attributes_value"
HINT:  Values larger than 1/3 of a buffer page cannot be indexed.
	Consider a function index of an MD5 hash of the value, or use full text indexing.
STATEMENT:  INSERT INTO "attributes" ("resource_id","key","value") VALUES ($1,$2,$3) RETURNING "attributes"."id"
ERROR:  index row size 4984 exceeds maximum 2712 for index "idx_attributes_value"
HINT:  Values larger than 1/3 of a buffer page cannot be indexed.
	Consider a function index of an MD5 hash of the value, or use full text indexing.
STATEMENT:  INSERT INTO "attributes" ("resource_id","key","value") VALUES ($1,$2,$3) RETURNING "attributes"."id"
ERROR:  index row size 5048 exceeds maximum 2712 for index "idx_attributes_value"

Had a look into the config and the code but couldn't figure out a fix yet.

[mvisonneau]

removing the index seems to solve the issue

terraboard=# DROP INDEX idx_attributes_value;
DROP INDEX

not sure if there is anything that can be configured around it's configuration :

terraboard/types/db.go

Line 56 in 9f1f536

Value string `gorm:"index" json:"value"`

[raphink]

Which version of postgresql is that?

[mvisonneau]

9.6.5-alpine https://hub.docker.com/_/postgres/

[mvisonneau]

mvisonneau@da3d8e1 fixes it but not quite an optimal solution

[raphink]

Yes indeed. Index was added here for search.