Re: unable to drop index because it does not exists

On 9/23/19 6:39 AM, Luca Ferrari wrote:
> Hi,
> running 11.5 I've got a partitioned table where I want to destroy an
> index (on a column that has nothing to do with the partition):
>
>
> respidb=# \d+ respi.root
>
> Column | Type |
> --------------+-----------------------------+-
> sen_id | integer |
> mis_flusso | integer |
> mis_tasso | integer |
> mis_velocita | integer |
> mis_ora | timestamp without time zone |
> pk | bigint |
> ts | timestamp without time zone |
> Partition key: LIST (date_part('year'::text, mis_ora))
> Indexes:
> "idx_root_sensore" btree (sen_id)
> "idx_ts" btree (ts)
> Partitions: respi.y2018 FOR VALUES IN ('2018'), PARTITIONED,
> respi.y2019 FOR VALUES IN ('2019'), PARTITIONED,
> respi.y2020 FOR VALUES IN ('2020'), PARTITIONED
>
> respidb=# drop index idx_root_sensore;
> ERROR: index "idx_root_sensore" does not exist
>
> Of course the index exists:
>
> SELECT oid, relname, relkind FROM pg_class WHERE relname = 'idx_root_sensore';
> -[ RECORD 1 ]-------------
> oid | 40950
> relname | idx_root_sensore
> relkind | I
>
>
> I already did a manual vacuum on the table.
> Any idea?

Try specifying the schema name:

drop index respi.idx_root_sensore;

--
Angular momentum makes the world go 'round.