Commit Graph

19 Commits

Author SHA1 Message Date
i-norden
c402e5c285 drop mh_keys and use cids for blockstore keys and linking aka revert to v0 ipfs blockstore format. in addition to saving space, this format is closer to the CAR format used in filecoin deals 2023-02-08 18:22:40 -06:00
i-norden
85bc243896 drop tx_data and log_data; data can still be accessed in referenced ipld blocks in public.blocks 2023-02-08 18:13:15 -06:00
i-norden
f67f03481b renaming some columns; remove log_root column 2023-02-08 18:08:08 -06:00
53461a0996 Add new indexes and functions to improve eth_getStorageAt performance. (#116)
* New indexes and functions to implement get_storage_at as a function.

* Update schema.sql.
2023-02-01 20:16:11 -06:00
i-norden
f3c58e39ca node_id => node_ids 2023-02-01 19:56:36 -06:00
i-norden
713f6a9208 add removed flag; embed value in storage_cids 2023-01-23 17:30:33 -06:00
i-norden
241bb281eb state_cids and storage_cids only index leaf nodes and related adjustments 2023-01-23 17:08:20 -06:00
i-norden
7f8247cb4f revert hash indexes; add eth probes tables 2022-08-30 12:05:51 -05:00
i-norden
2bf5c82150 replace btree with hash index where it makes sense 2022-08-08 12:40:47 -05:00
i-norden
0856168b92 misc adjustments 2022-08-08 11:06:00 -05:00
prathamesh0
b59505eab2
Add block hash to primary keys in transactions, receipts and logs tables (#100)
* Add block hash to primary keys in transactions, receipts and logs tables

* Add block hash in postgraphile triggers for transactions, receipts and logs tables

* Make indexes on transaction cid and mh_key non-unique
2022-07-07 16:21:06 +05:30
i-norden
673ae7b265 fixes, PKs need to include partition key 2022-04-12 21:56:27 -05:00
i-norden
caf99f7194 use state leaf key in storage_cids FK 2022-04-12 21:11:36 -05:00
i-norden
da8d0af6df updates order of columns in compound PKs, update indexes e.g. we don't need a btree index on a column if it is the first column in the compound PK index but we do need a btree index for the later columns in a compound PK (searches on first column of a compound index are just as fast as searches on a btree index for that column alone, but searches on the 2nd or 3rd column in a compound index are significantly slower than on dedicated indexes) 2022-03-28 18:22:57 -05:00
i-norden
1dc12460dc denormalize tables by block_number so that we can partition all tables by block_number for purposes of sharding 2022-03-15 16:06:13 -05:00
i-norden
f59582ab42 indexes on receipt cid and mh_key should not be unique, as it is possible (but improbable) that two receipts can be identical 2022-02-16 14:11:23 -06:00
i-norden
9e7e0377a5 split pk application into two parts 2022-01-26 13:01:41 -06:00
i-norden
b8f0c194b8 remove pks; migrations to add back pks and fks 2021-11-18 17:46:50 -06:00
i-norden
7cc9201a94 new natural primary key scheme 2021-11-14 18:23:26 -06:00