GEP Grid Upgrade to Version Alpha-5 (A5) (CLOSED)

After a few weeks of testing we think we’re ready to go full Monty on the grid and upgrade it to Alpha-5.

This entails:

  • Got smarter in selecting Planetary network endpoints/peers
  • Kernel minor bump. No need to reboot, but will be active after a reboot
  • Some fixes that erroneously deselected ssd as fast storage when connected to a raid controller
  • VMs are now configured through cloud-init (giving access to generic cloud-images for users of the grid)
  • Vm logs can get streamed to a remote Redis an/or WS endoint, so users can verify logs of the started VM
  • Changes necessary to properly interact with the new features in TFChain
  • Azmy says: ‘many bugs’. Don’t believe him, just some small cosmetics. But if you’re That guy:


  • GraphQL frontend to the chain gets a major upgrade
  • Fixes to billing of smart contracts
  • General stability improvements


  • Dedicated node support. The ability to rent a node as a whole, any deployment on that node is incorporated into the price of renting a node). Farmers can mark their farm as dedicated per request. (best to do that request via the Forum, there is no automated way for that yet)

As we’re relying on the Community’s Infinite Wisdom, we ask for It’s invaluable input with a poll:

  • Go for it
  • Please wait

0 voters


Nice, love seeing this first step towards managing Zos upgrades by DAO :slight_smile:


Hi Jan, I might not be “that guy” but I did click the link and it ended up on a 'Page not found". @delandtj can you have a look. I think you mean this one, but not sure:

Something happened. I accidentally released because I wanted to fix a typo in Azmy’s changelist.
That triggered an unwanted deployment, while the chain wasn’t ready.
As that was an unaccounted path, things ‘happened’ that have shown we’re not ready yet.
We’re investigating what ‘happened’ specifically, to ascertain that other ‘things’ won’t ‘happen’ too on mainnet.
SO the GEP, although called, will be extended to next wednesday at least, still need to see with team here.

I don’t believe this shows that we are ‘not’ ready yet. the problem happened on testnet where it was upgraded, then downgraded, then upgraded again. and zos is not designed to handle downgrade.

1 Like

keep the date then ?

As @azmy says it’s not a sign of not being ready I would say keep the date!

I definitely say deploy, the only action guaranteed to not improve things is inaction.

1 Like

Will kernel upgrades require a reboot in the future?

Terraform users, remember to update you grid provider with:

terraform init -upgrade

You need: threefoldtech/grid v0.1.31

1 Like