> Sankar P writes:
> > I spoke too soon. While this worked fine when there were no indexes
> > and finished within 10 minutes, with GIN index on the jsonb column, it
> > is taking hours and still not completing.
>
> There's the generic advice that building an index after-the-fact
> is often cheape
Sankar P writes:
> I spoke too soon. While this worked fine when there were no indexes
> and finished within 10 minutes, with GIN index on the jsonb column, it
> is taking hours and still not completing.
There's the generic advice that building an index after-the-fact
is often cheaper than updati
>
> I spoke too soon. While this worked fine when there were no indexes
> and finished within 10 minutes, with GIN index on the jsonb column, it
> is taking hours and still not completing.
>
It is always recommended to create indexes AFTER loading data. Sometimes it
can be faster to drop all index
> > It looks like plain old COPY would do this just fine, along the lines
> > of (in psql)
> >
> > \copy myTable(content) from 'myfile.ndjson'
I spoke too soon. While this worked fine when there were no indexes
and finished within 10 minutes, with GIN index on the jsonb column, it
is taking hours
> It looks like plain old COPY would do this just fine, along the lines
> of (in psql)
>
> \copy myTable(content) from 'myfile.ndjson'
>
Indeed. Thanks.
--
Sankar P
http://psankar.blogspot.com
Sankar P writes:
> I have a .ndjson file. It is a new-line-delimited JSON file. It is
> about 10GB and has about 100,000 records.
> Some sample records:
> { "key11": "value11", "key12": [ "value12.1", "value12.2"], "key13": {
> "k111": "v111" } } \n\r
> { "key21": "value21", "key22": [ "value22.1"
Hi
I have a .ndjson file. It is a new-line-delimited JSON file. It is
about 10GB and has about 100,000 records.
Some sample records:
```
{ "key11": "value11", "key12": [ "value12.1", "value12.2"], "key13": {
"k111": "v111" } } \n\r
{ "key21": "value21", "key22": [ "value22.1", "value22.2"] }
```
