Nodes have error "exit status 2"

[MorgainFriedman]

None of the nodes start and stay running. Running on Umbrel

[bochaco]

Hi @MorgainFriedman , can you please confirm which version of Formicaio you are currently running and seeing that issue with? also, please let me know which antnode binary version is showing that behavior.

Are you installing Formicaio from the official UmbrelOS app store or from the Formicaio community app store at https://github.com/bochaco/formicaio-app-store ?
I would recommend installing from the latter where I published the very latest version of Formicaio and it is not yet published on the official UmbrelOS app store.

[MorgainFriedman]

From the official app store. 0.8.4 version .

How do I find "antnode binary version"?

[bochaco]

How do I find "antnode binary version"?

In the nodes page, expand the node's card, it'll show the node binary version.

From the official app store. 0.8.4 version .

Install Formicaio from the Formicaio community app store at https://github.com/bochaco/formicaio-app-store (look at the video in that repo for how to do it), v0.8.7 is the latest and should be compatible with latest antnode, let me know otherwise.

[MorgainFriedman]

So I installed the community version. All nodes are active, but it says this in the logs:

2026-08-13T21:59:14.348207Z WARN saorsa_core::reachability::session: relay acquisition: no direct-addressable candidates in routing table
2026-08-13T21:59:14.348241Z WARN saorsa_core::reachability::driver: driver: acquisition failed, clearing canary exclusions and entering backoff reason="no direct-addressable candidates in routing table" rejected_relayers=0
2026-08-13T21:59:14.843408Z WARN saorsa_transport::candidate_discovery: Local interface scan timeout for Local, proceeding with available candidates
2026-08-13T21:59:14.843430Z INFO saorsa_transport::candidate_discovery: Discovery completed (timeout) for Local: 8 candidates
2026-08-13T21:59:46.166498Z WARN saorsa_core::reachability::session: relay acquisition: no direct-addressable candidates in routing table
2026-08-13T21:59:46.166525Z WARN saorsa_core::reachability::driver: driver: acquisition failed, clearing canary exclusions and entering backoff reason="no direct-addressable candidates in routing table" rejected_relayers=0
2026-08-13T22:00:12.863003Z INFO ant_node::replication: Replication bootstrap state snapshot drained=true pending_peer_requests=0 pending_keys=0 capacity_rejected_sources=0 is_bootstrapping=false snapshot_state="healthy" uptime_secs=60
2026-08-13T22:00:12.863049Z INFO ant_node::replication: Replication bootstrap diagnostics reached healthy state; stopping periodic snapshots
2026-08-13T22:00:46.521568Z WARN saorsa_core::reachability::session: relay acquisition: no direct-addressable candidates in routing table
2026-08-13T22:00:46.521604Z WARN saorsa_core::reachability::driver: driver: acquisition failed, clearing canary exclusions and entering backoff reason="no direct-addressable candidates in routing table" rejected_relayers=0

[bochaco]

I'll be double checking if there were breaking changes in the last few antnode releases and I'll be sharing an update, along with a new Formicaio version if necessary.

[bochaco]

There doesn't seem to be any breaking change introduced in antnode that could make Formicaio v0.8.7 incompatible with latest antnode v0.17.1.

I've just tried it myself, and after only a couple of minutes, it already shows some chunks have been downloaded, which makes me think it is connecting correctly to the network (I don't get those warning logs though).

It sounds it may be more related to your own specific network connectivity and/or setup.
If you still have issues, or those warnings are suspiciously making your node misbehave, I'd advise you to report it to ant-node repo https://github.com/WithAutonomi/ant-node / Autonomi and they should be able to share more details about them.

[MorgainFriedman]

Do I need MCP server? It says inactive

[MorgainFriedman]

How many nodes should be run? I don't see any information about that.

[bochaco]

Do I need MCP server? It says inactive

No you don't, unless you have an AI agent you want it to connect to Formicaio, but that's not supported on UmbrelOS deployment, you'd need to run Formicaio natively, you have instructions here: https://github.com/bochaco/formicaio#mcp-server-for-integration-with-ai-agents

How many nodes should be run?

As many as you want and according to your machine's capacity, the more you run the more chances you have to receive more rewards. I'd advise you to check Autonomi's website though.

[MorgainFriedman]

I have ollama local connected to ai agent its connected successfully. I can't find where it says how many nodes for however many specs

[MorgainFriedman]

A day later and no chunks are being stored

[bochaco]

A day later and no chunks are being stored

It may be then that your node is not connecting to the rest of the network, it could be your own specific network connectivity and/or setup as per the warnings messages you are seeing. You'd need to report it to ant-node repo https://github.com/WithAutonomi/ant-node / Autonomi team.

[MorgainFriedman]

Unfortunately ive done so but havent gotten a response