docs(tvix/TODO): extend NAR rendering prefetching idea
With the seekable NAR renderer, figuring out the next few blobs to render became possible. Change-Id: I1214302f88e6f9aba74227f84df0f964d587baf2 Reviewed-on: https://cl.tvl.fyi/c/depot/+/12652 Tested-by: BuildkiteCI Reviewed-by: edef <edef@edef.eu> Autosubmit: flokli <flokli@flokli.de>
This commit is contained in:
parent
1c80bc4b5b
commit
4ebb8f42ed
1 changed files with 7 additions and 2 deletions
|
|
@ -75,8 +75,13 @@ correctness:
|
||||||
This is somewhat the "spiritual counterpart" to our sequential ingestion
|
This is somewhat the "spiritual counterpart" to our sequential ingestion
|
||||||
code (`ConcurrentBlobUploader`, used by `ingest_nar`), which keeps
|
code (`ConcurrentBlobUploader`, used by `ingest_nar`), which keeps
|
||||||
"some amount of outgoing bytes" in memory.
|
"some amount of outgoing bytes" in memory.
|
||||||
This is somewhat blocked until the {Chunk/Blob}Service split is done, as then
|
Our seekable NAR AsyncRead implementation already removes most complexity in
|
||||||
prefetching would only be a matter of adding it into the one `BlobReader`.
|
rendering everything between blobs.
|
||||||
|
It should be possible to extend / write a wrapped version of it that
|
||||||
|
prefetches a configurable sliding window of blobs.
|
||||||
|
Per-blob prefetching itself is somewhat blocked until the {Chunk/Blob}Service
|
||||||
|
split is done, as then prefetching there would only be a matter of adding it
|
||||||
|
into the one `BlobReader`.
|
||||||
|
|
||||||
### Error cleanup
|
### Error cleanup
|
||||||
- Currently, all services use tvix_castore::Error, which only has two kinds
|
- Currently, all services use tvix_castore::Error, which only has two kinds
|
||||||
|
|
|
||||||
Loading…
Add table
Add a link
Reference in a new issue