Following on from a discussion in discord, a small tweak to make the required key length for a HMAC key more obvious

* added feedback on migration commands (close hasura#1410) (hasura#1931)

* firebase2graphql (v0.0.2): support array types in firebase json (close hasura#1829) (hasura#1900)

* improve console error page (close hasura#272) (hasura#1890)

* add nextjs 8 jwt sample app (hasura#1943)

* persist queries in GraphiQL (closes hasura#1894) (hasura#1903)

* add realtime chat sample app with vue (hasura#1885)

* fix update_event_trigger query not persisting the changes (hasura#1950)

fix hasura#1949

* mention required key length for HMAC encryption to user (hasura#1956)

Following on from a discussion in discord, a small tweak to make the required key length for a HMAC key more obvious

* fix input argument types for custom SQL functions (close hasura#1952) (hasura#1953)

* run graphql tests on both http and websocket; add parallelism (close hasura#1868) (hasura#1921)

Examples 
1)  `
pytest --hge-urls "http://127.0.0.1:8080" --pg-urls "postgresql://admin@127.0.0.1:5432/hge_tests" -vv
`
2)  `pytest --hge-urls "http://127.0.0.1:8080"   "http://127.0.0.1:8081" --pg-urls "postgresql://admin@127.0.0.1:5432/hge_tests"  "postgresql://admin@127.0.0.1:5432/hge_tests2" -vv
`
### Solution and Design
<!-- How is this issue solved/fixed? What is the design? -->
<!-- It's better if we elaborate -->
#### Reducing execution time of tests
- The Schema setup and teardown, which were earlier done per test method, usually takes around 1 sec. 
- For mutations, the model has now been changed to only do schema setup and teardown once per test class.
-  A data setup and teardown will be done once per test instead (usually takes ~10ms).
- For the test class to get this behaviour, one can can extend the class `DefaultTestMutations`. 
    - The function  `dir()` should be define which returns the location of the configuration folder.
    - Inside the configuration folder, there should be 
        - Files `<conf_dir>/schema_setup.yaml` and `<conf_dir>/schema_teardown.yaml`, which has the metadata query executed during schema setup and teardown respectively
        - Files named `<conf_dir>/values_setup.yaml` and `<conf_dir>/values_teardown.yaml`. These files are executed to setup and remove data from the tables respectively. 

#### Running Graphql queries on both http and websockets
- Each GraphQL query/mutation is run on the both HTTP and websocket protocols
- Pytests test parameterisation is used to achieve this
- The errors over websockets are slightly different from that on HTTP
   - The code takes care of converting the errors in HTTP to errors in websockets

#### Parallel executation of tests.
- The plugin pytest-xdist helps in running tests on parallel workers.
- We are using this plugin to group tests by file and run on different workers.
- Parallel test worker processes operate on separate postgres databases(and separate graphql-engines connected to these databases). Thus tests on one worker will not affect the tests on the other worker.
- With two workers, this decreases execution times by half, as the tests on event triggers usually takes a long time, but does not consume much CPU.

* update manifests to v1.0.0-alpha42

* update regex to delete event triggers in edit-pg-dump (close hasura#1959) (hasura#1960)

* support composite foreign keys and primary keys through console (close hasura#1707) (hasura#1854)

* edit-pg-dump: fix a missing semi-colon

* fix styles in browse rows, relationship and permission tables (hasura#1969)

* update console permissions ui (hasura#1924)

* Make permissions sections as collapsibles with tooltips
* Remove 'use same permission as select' type options from row permissions section
* Added clone section to apply same permission to any table-role-action
* Disable other subsections till row permissions are set

* re-add same as other action options in permissions (hasura#1970)

* fix excessive memory consumption with bulk queries (hasura#1972)

Use 'Strict' State monad instead of 'Lazy' to avoid unevaluated memory thunks

* add placeholder for bool type fields while inserting rows (hasura#1962)

* Use UTF-8 encoding for stack hpc report generation (hasura#1974)

* fix save function in console's retry config editor (fix hasura#1975) (hasura#1976)

* Wip

* add scroll to literal blocks in docs (close hasura#1954) (hasura#1978)

* remote relationship ui v1

* wip

* ui for creating remote relationships

* wip

* ui fixes

* ui fixes

* functional ui

* namespace ui wip

* add functionality for dropping relationship

* show exact error in notification

* fix state inconsistencies

* update reldef for namespaced fields

* lint

Following on from a discussion in discord, a small tweak to make the required key length for a HMAC key more obvious

These changes were tested/profiled against

```
$ cargo bench --bench benchmarks --  --measurement-time 20 'build example/big_pretty'
```

while taking samples of memory usage a couple times per second.

Compared to my original baseline, this **improves peak memory usage by
~60% and latency by ~20%**. (FYI after the final rebase memory usage is
still good, latency is somewhat worse than before the rebase but still
better than my original baseline)

## IYI Global Allocator observations

When v3 memory issues first came up I discovered [this
issue](purpleprotocol/mimalloc_rust#141) and
suggested we could try `MIMALLOC_PURGE_DELAY=0` in production as a cheap
thing to try. Since I needed to temporarily disable mimalloc to use
heaptrack anyway, I did a few tests; these are the same commit (probably
the first change here):

| variant | latency (s) |
| -------- | --------------------- |
mimalloc (default): | 4.15
no mimalloc: | 5.31
mimalloc w/ `MIMALLOC_PURGE_DECOMMITS=0`: |  3.70
mimalloc w/ `MIMALLOC_PURGE_DELAY=0`: | 5.23

Observations:
- `MIMALLOC_PURGE_DELAY=0` might negate any performance benefits of
mimalloc
- `MIMALLOC_PURGE_DECOMMITS=0` (using `MADV_FREE` to lazily free memory
back to OS) was a significant optimization (but still not something I
think we or anyone can use)

With `MIMALLOC_PURGE_DELAY=0` we certainly saw memory freed in a much
more fine-grained way throughout the phases of each benchmark iteration.

V3_GIT_ORIGIN_REV_ID: 23fe5f1200dea9845a337f8880f4a38189aa1b87