Conversation
frrist
commented
Jan 19, 2023
- related to Lotus daemon being killed and causing websocket errors lotus#10050
|
I'm having some issues with this PR. This is the The daemon is kept running but the walk is not completing. The list command outputs this: [
{
"ID": 1,
"Name": "walk_1674391696",
"Type": "walk",
"Error": "",
"Tasks": [
"block_header",
"block_parent",
"drand_block_entrie",
"data_cap_balance",
"miner_beneficiary",
"miner_sector_deal",
"miner_sector_infos_v7",
"miner_sector_infos",
"miner_sector_post",
"miner_pre_commit_info",
"miner_sector_event",
"miner_current_deadline_info",
"miner_fee_debt",
"miner_locked_fund",
"miner_info",
"market_deal_proposal",
"market_deal_state",
"message",
"block_message",
"receipt",
"message_gas_economy",
"parsed_message",
"internal_messages",
"internal_parsed_messages",
"vm_messages",
"multisig_transaction",
"chain_power",
"power_actor_claim",
"chain_reward",
"actor",
"actor_state",
"id_addresses",
"derived_gas_outputs",
"chain_economics",
"chain_consensus",
"multisig_approvals",
"verified_registry_verifier",
"verified_registry_verified_client"
],
"Running": true,
"RestartOnFailure": false,
"RestartOnCompletion": false,
"RestartDelay": 0,
"Params": {
"maxHeight": "2399050",
"minHeight": "2396162",
"storage": "CSV",
"window": "0s"
},
"StartedAt": "2023-01-22T12:48:16.859521259Z",
"EndedAt": "0001-01-01T00:00:00Z"
}
] |
|
@davidgasquez is it failing for all epochs or just some? |
|
I think is just some (will check tomorrow when I'm back). That said, it happens in most of them and should be easy to reproduce in any recent partial archive. |
|
I've used The There is also a bunch of empty files: Finally, in That said, the walk seems to continue without problems after these websocket errors. |
1 similar comment
|
I've used The There is also a bunch of empty files: Finally, in That said, the walk seems to continue without problems after these websocket errors. |
- prevents panic
|
Tested things locally and seems its working fine. I've got some errors but I think they're unrelated with this PR: Will send some jobs and keep an eye on things. |
|
Heads up that I just got the first |
|
@frrist do you have any ideas about why it might be panicking in the first place? |
|
@hsanjuan current theory is filecoin-project/filet#22 (comment) |
|
The "maybe there are blocks missing in a datastore" doesn't fit the randomness that seems to affect this issue. |
|
@davidgasquez can you point me to the github issue for the websocket bug? |
There is one higher level in Filet and another in Lotus. |
|
@frrist should we merge? |