This Week in Lotus - Week 38
Posted September 23, 2022 by ‐ 2 min read
👋 Hey everyone and welcome to This Week in Lotus!
The first release candidate for Lotus v1.17.2
was released this week and is currently under testing by the SPX group. This version enables the sealing-as-a-service API in Lotus. This will allow builders to create sealing services on top of Lotus. If you have a special interest in that subject please check out the #fil-sealing-as-a-service
channel on Slack. In addition to the SaaS-feature, the v1.17.2 release also includes a lot of bug fixes, enhancements and new features. You can read the full releaselog here.
📣 Please note that Lotus v1.17.2 will require a Go-version of v1.18.1 or higher!
🐛 fixes:
- A bug where running
lotus-miner sectors update-state
on a sector in theReplicaUpdateFailed
state caused thelotus-miner
to crash. A fix to avoid the panic has been merged, and is also included in the v1.17.2-rc1 version. - The FFI has been updated to fix an issue where sectors on read-only storage was skipped during windowPoSt. (This fix is also in v1.17.2-rc1)
- A fix to revert the
lotus-miner init
default size back to 32GiB has been merged. - A price-per-byte calculation fix for retrievals to make the calculation correct has been merged.
- A regression on the master-branch caused the
lotus state compute-state
command to return too much information for it to unmarshal. The regression was fixed and is now in the master branch.
Protocol development:
A lot of work on implementing and integrating FIP0045 - De-couple verified registry from markets
has happened this week:
- On the builtin-actors side of the code, you can see all the work items finished this week here.
- On the Lotus side:
- A PR to integrate the builtin-actors changes for FIP-0045 has been opened.
- Integrating DataCap actors into Lotus
N.B for Storage Providers: In v1.17.2 the default PropagationDelay
has been raised from 6 seconds -> 10 seconds, and you can tune this yourself with an environment variable. That means you will now wait for 10 seconds for other blocks to arrive from the network before computing a winningPoSt (if eligible). In your lotus-miner logs that means you will see this "baseDeltaSeconds": 10
as default. You can read the full post about why this was raised here.
That´s it for the week! Have a great weekend! ☀️