You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: CHANGELOG.md
+12-1Lines changed: 12 additions & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -16,6 +16,10 @@ If you were at `firehose-core` version `1.0.0` and are bumping to `1.1.0`, you s
16
16
17
17
### Substreams
18
18
19
+
#### Reconnection time
20
+
21
+
* Added flag `substreams-tier1-quicksave-store` to enable quicksave of stores on tier1, allowing for a fast reconnection of clients using stores.
22
+
19
23
#### Capacity Management
20
24
21
25
* Integrated the `GlobalRequestPool` service in the `Tier1App` to manage global requests pooling.
@@ -34,10 +38,17 @@ If you were at `firehose-core` version `1.0.0` and are bumping to `1.1.0`, you s
34
38
35
39
#### Performance
36
40
37
-
* Fixed a regression since "v1.7.3" where the SkipEmptyOutput instruction was ignored in substreams mappers
38
41
* Add shared cache for tier1 execution near HEAD, to prevent multiple tier1 instances from reprocessing the same module on the same block when it comes in (ex: foundational modules)
39
42
* Improved fetching of state caches on tier1 requests to speed up "time to first data"
40
43
44
+
* Rust modules will now be executed with `wasmtime` by default instead of `wazero`.
45
+
- Prevents the whole server from stalling in certain memory-intensive operations in wazero.
46
+
- Speed improvement: cuts the execution time in half in some circumstances.
47
+
- Wazero is still used for modules with `wbindgen` and modules compiled with `tinygo`.
48
+
- Set env var `SUBSTREAMS_WASM_RUNTIME=wazero` to revert to previous behavior.
49
+
50
+
* Fixed a regression since "v1.7.3" where the SkipEmptyOutput instruction was ignored in substreams mappers
51
+
41
52
### Tools
42
53
43
54
* make 'compare-blocks' command support one-blocks stores as well as merged-blocks
This is useful to prevent the tier1 from being overwhelmed by too many requests, most client auto-reconnects on 'Unavailable' code
74
74
so they should end up on another tier1 instance, assuming you have proper auto-scaling of the number of instances available.
75
75
`))
76
+
cmd.Flags().String("substreams-tier1-quicksave-store", "", "If enabled, substreams will use this store to put 'quicksave' data when shutting down while running requests with 'stores'. Use this flag with a non-zero --common-system-shutdown-signal-delay")
76
77
cmd.Flags().String("substreams-tier1-global-worker-pool-address", "", "Address of the global worker pool to use for the substreams tier1. (disabled if empty)")
77
78
cmd.Flags().String("substreams-tier1-global-request-pool-address", "", "Address of the global worker pool to use for the substreams tier1. (disabled if empty)")
78
79
cmd.Flags().Duration("substreams-tier1-global-worker-pool-keep-alive-delay", 25*time.Second, "Delay between two keep alive call to the global worker pool. Default is 25s")
0 commit comments