ffe776b70d
AccountWriteFilter generic router to process account writes for different modules in the same process
2023-02-07 17:22:21 +09:00
b9afc2c200
Merge branch 'v4' into max/crank
2023-02-05 10:53:02 +09:00
4b75b88a98
cargo fmt
2023-02-03 11:49:22 +00:00
5c8ebc53b1
Fills updates
...
* Unify fill event schema
* Change fill updates json
* Convert all native values to ui
* Add fills postgres target
2023-02-03 11:39:53 +00:00
31149f7e0d
Merge remote-tracking branch 'origin/v4' into max/crank
2023-02-03 16:22:35 +09:00
5e78f008f6
create crank skeleton
2023-02-03 16:15:58 +09:00
3dc7ae1246
* Add chaindata metrics
...
* jemalloc for fills
* Reenable dropped fill processing
* Add gMA snapshot support
* Tidy up serum orderbook change detection
* cargo fmt
2023-01-20 14:52:01 +00:00
49bcdeee77
Use new geyser plugin, add orderbook service
2022-12-16 11:00:42 +00:00
2881378b7f
Refactor metrics
2022-10-07 11:44:53 +01:00
b41c094aba
merge fill service
2022-03-28 20:01:16 +02:00
cdf6d54156
Config: Separate out source config
...
We want to be able to not have a postgres target
2022-03-23 08:51:17 +01:00
3beaac671d
Add memory_target and chain_data
...
This will be useful for storing account data in memory.
2022-03-23 08:19:19 +01:00
42f07d58a6
Keep slot, write_version etc as u64, convert only for postgres
...
This makes the conversion to i64 something that's local to the postgres
target, and doesn't affect unrelated code.
2022-03-23 08:18:02 +01:00
f25c49d983
Configurable max account write queue size
2022-03-16 11:04:30 +01:00
80423fdac9
Track "is_selected" to denote closed/reused accounts
2022-01-28 10:12:01 +01:00
dc80d8266c
Process postgres cleanup jobs separately
...
Executing them for every rooted slot may overwhelm the database.
2022-01-20 14:49:51 +01:00
7d8fd45de4
Add support for updating a monitoring table
...
That way connector interruptions can be seen directly form the target
database.
2022-01-06 12:54:35 +01:00
5b5eaba4ff
Pipeline SQL inserts for better performance
...
Previously insert speed was primarily limited by roundtrip time to the
database since they were fully sequential in each connection. Now a full
batch of inserts is done per connection in parallel.
This could still be improved a lot. Currently each connection will fully
wait for all inserts to be done in the current batch before starting
another one.
2022-01-04 10:15:57 +01:00
d5395a3b75
Add config flag for deleting old data
...
Keeping it is very useful for testing.
2022-01-03 10:36:11 +01:00
17d6787bf1
Add TLS domain name argument
2021-11-30 17:15:41 +01:00
b65815e4d4
gRPC: Optionally use TLS
2021-11-18 14:21:17 +01:00
dde006243b
Minor clippy fixes
2021-11-15 15:37:29 +01:00
8f6bd7bc22
Postgres: Allow parallel slot update connections
...
It would be nicer if many of the writes could happen on the same
connection. Separate connections are only necessary if transactions are
used.
2021-11-15 15:28:18 +01:00
343b5408bc
Postgres: Add option to allow invalid certificates
2021-11-14 13:14:19 +01:00
d889b5ff5b
Move postgres numeric support to separate file
2021-11-09 14:32:58 +01:00
3a0e2535f0
Use anyhow::Result over Result<T, anyhow::Error>
2021-11-09 14:27:09 +01:00
4d5f0c5d10
Basic metrics
2021-11-09 09:00:00 +01:00
029cbd2fb5
Allow multiple grpc sources
2021-11-08 15:48:50 +01:00
11f4d6bd67
Use separate pubkey table
2021-11-08 14:10:48 +01:00
a17f142875
Use enum for slot status
2021-11-08 12:44:01 +01:00
47f8bb823d
Remove unnecessary trailing backslashes from multiline string literals
2021-11-08 11:59:57 +01:00
0424c3c9d6
grpc configurability
2021-11-08 11:42:22 +01:00
d366581826
Make more postgres settings configurable
2021-11-08 11:30:17 +01:00
d9cb30c2b3
Make parallel postgres connections configurable
2021-11-08 10:05:17 +01:00
cbe9e16b62
Configurable program id
2021-11-08 09:57:56 +01:00
e67dd6c235
Split into connector library, and raw and mango binaries
2021-11-08 09:45:41 +01:00
Maximilian Schneider
Riordan Panayides
Christian Kamm
Christian Kamm