-
Notifications
You must be signed in to change notification settings - Fork 4
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Generate both ERA and ERA1 files for Sepolia testnet #189
Comments
If Geth re-sync is necessary, it should be done one node at a time, preferably with the respective BN using a different Geth node. |
The simplest way to check Geth and Beacon Node sync state is using the RPC and REST APIs respectively:
|
The second way to check Geth and Beacon Node sync stage is using metrics in Grafana: Unfortunately on the Sepolia host the Geth containers do not have an accompanying
You don't have to deploy them, but they can be useful when long-term re-sync processes run. You can see how it's set up here: infra-nimbus/ansible/group_vars/nimbus.holesky.yml Lines 31 to 35 in 4555e37
infra-nimbus/ansible/holesky.yml Lines 58 to 65 in 4555e37
|
The way to view Beacon Node sync progress is also to use Grafana: These metrics require no additional setup. |
These metrics are of course available on the host itself by using the correct metrics service port:
|
If you want to check Geth docker container configs you need to look at their Docker Compose files:
And if you want to check Beacon Node systemd service configs you need to look at the service and the config file:
|
The main configuration files for any Nimbus fleets are in a few locations, here's a list for Sepolia:
|
Logging size after changing sync mode of libp2p sepolia to snap sync manually ( not via ansible ).
Will check in a bit if size is purged or not. |
I ( with the help of Jakub) generated the ERA1 files like this :
|
@advaita-saha : could you please take a look at these generated ERA 1 files and let us know if we missed something?
|
Since it seems like switching To do a Geth resync we need to:
This way we can avoid unnecessary downtime for the validators. |
@siddarthkay the Era1 files looks good to me |
@siddarthkay you can add hosting for ERA1 files the same way we do it for ERA files: infra-nimbus/ansible/host_vars/linux-01.ih-eu-mda1.nimbus.sepolia.yml Lines 30 to 44 in 2275872
Change: 458652e I probably would just add an extra domain, for example |
I manually re-synced
I'll continue to re-sync other nodes tomorrow |
I've manually re-synced
|
I've re-synced Total space cleaned up is -> 591 + 599 + 591 + 591 - (353 + 361 + 362 + 361) = 935 GB
|
Now it's time to resync a Beacon Node right? |
@jakubgs : correct, I'll first move era and era1 files to
|
I've moved
|
I've re-synced stable beacon node and it took roughly 3 days to sync. The fully synced node takes ~ 300 GB of space.
Now time to generate missing ERA files. |
I modified I then had to :
And then I could see that missing ERA file generation was successful
|
I've set history back to prune for sepolia stable beacon node and restarted the service for the config change to take effect.
|
re-syncing completed last night. ![]() and we've reclaimed the space.
Thanks @jakubgs, this task gave me good insights and made me a bit comfortable with the nimbus fleet. |
the archive at https://sepolia.era1.nimbus.team/ is incomplete - there should be 183 files cc @siddarthkay |
@arnetheduck : https://sepolia.era1.nimbus.team/ should be complete now, confirmed with @advaita-saha I regenerated ERA1 files like this
|
There are two types of chain state archive files:
Currently we are hosting ERA files under https://sepolia.era.nimbus.team/ but some of them are missing:
Unfortunately our Beacon Nodes are running with
history
mode set toprune
, which means they only retain last 6 months of data. In order to re-generate this new data we need to purge BN data, and switch thehistory
mode toarchive
, and let it re-sync. The current available storage does not make that possible since a fully synced BN might take up up to 1 TB:For this reason we need to perform these steps to get what we want hopefully without purchasing extra storage:
snap
mode with--syncmode
to reclaim space./data/era
files to/docker
to recover space necessary for BN re-sync.stable
BN with history retention set toarchive
using reclaimed space.history
configuration flag which can beprune
orarchive
.nimbus-era-files-update.service
.prune
history mode to reclaim space.Once done the ERA1 files should also be hosted. Probably under: https://sepolia.era1.nimbus.team/
This is a low priority task.
The text was updated successfully, but these errors were encountered: