Invalid SSZ: trying to read more bytes than available

[chrishobcroft]

I have the following running on a machine :

• geth - fully synced with Ephemery

• teku - talking to geth, and synced with Ephemery, with --rest-api-enabled=true

• checkpointz - with teku upstream on http://localhost:5051

• geth2 - ready to work with teku2 to sync, with non-default port numbers im order to avoid port clashing

I am trying to run teku2 with:

./teku/build/install/teku/bin/teku     --network ~/testnet-all/config.yaml     --initial-state=http://localhost:5555     --data-path "datadir-teku-2"  --data-storage-mode archive --rest-api-enabled=true  --ee-endpoint http://localhost:8552     --ee-jwt-secret-file "/tmp/jwtsecret"     --log-destination console     --p2p-discovery-bootnodes "enr:-Iq4QGdecZrX_n9aWd0T0r3w9aVTnjbrO9nyDSMunKwxh6M3fkTuDamFtcu4Ulgq82WA7V10XJJJgDFUqrYEGAonNGiGAYSHF-YRgmlkgnY0gmlwhIjzHgyJc2VjcDI1NmsxoQJDyix-IHa_mVwLBEN9NeG8I-RUjNQK_MGxk9OqRQUAtIN1ZHCCIyg,enr:-Ly4QOHA0l0oGZ_usY-LbPlGlchutlXMcON_IDPpm_6VGpK1SqVxIKzWqa7HFddwcZV-HPwcOTKZsVGQIxKk_guF-o4Bh2F0dG5ldHOIAAAAAAAAAACEZXRoMpAJdJ_LMAAQGf__________gmlkgnY0gmlwhIe1jKiJc2VjcDI1NmsxoQPpDPi6_JtqjVqPKKmMnOU3M_hhK10zWtOqtJWgUlVw-IhzeW5jbmV0cwCDdGNwgiNQg3VkcIIjUA,enr:-Jq4QLPdytrxHEAHiTbd8jh6Yt8Ma7xD2V46U-rWYEtN13s1EJQCKhbSEN1NVaBbp0dB-6rBBxEEBXHadNpuz3jxToUBhGV0aDKQCXSfyzAAEBn__________4JpZIJ2NIJpcISI8x4MiXNlY3AyNTZrMaEDBJTLsTGGR6Ja1ezBFFYm3BgvwoB0U91jTtBfnl_x296DdWRwgiMp" --p2p-port=9001 --rest-api-port=5052

i.e. pulling --initial-state from checkpointz running on localhost:5555.

When I run teku2, it fails with:

Invalid SSZ: trying to read more bytes than available

It seems that teku finds checkpointz (running teku2 without checkpointz running gives a different error), but cannot read from it.

[chrishobcroft]

This is a screenshot of my checkpointz running:

[samcm]

I think this might be because your checkpointz instance is running in light mode, i.e. it won't fetch/server any state requests. If you toggle over to full mode with a config like this it might just work 🙏

checkpointz:
  mode: full

[chrishobcroft]

OK, this has resolved the last issue... now I am getting:

The fork in loaded state does not match fork at the epoch from ForkSchedule. Please check that network in configuration matches the loaded state.

Here is checkpointz's config.yaml:

And this is how I am starting the teku that runs http://localhost:5051 :

where the --network ~/testnet-all/config.yaml is from here.

If there is not anything obvious that I am doing wrong please feel free to close this issue!

[samcm]

I can recreate what you're seeing with Teku, but Lighthouse seems to work which is weird. I don't really know enough about SSZ to know what this is off the top of my head, so will need to dig in to it a little bit sorry!