Can't deploy programs to the localnet validator started by workbench #26
Comments
|
mmm, ok, not worknet - I'm getting the same problem from https://github.com/solana-labs/example-helloworld/blob/master/README.md |
SvenDowideit
changed the title
|
I wonder if the local validator did go pop - there's not much in my logs, but it does end in nope, restarting it didn't do it. |
|
local validators in docker are based on the solana CI images, which might have non-standard config, i assumed they would be OK, but apparently not for real use. currently experimenting with building a custom solana docker image to see if it will work better |
|
blech, even after trying building Solana in Docker from I wonder what is going on and how to debug. solana-labs/solana#18033 seems related, but you also couldn't get it to work on your x86_64 linux box right? |
|
maybe - but i didn't actually poke it hard. |
|
so this works fine within the container itself (after installing |
|
will have to double check that these are available, i wonder what the high ports are the ports i have open are: this is on D4M |
|
ok this is weird, |
|
hm i think it does but i have no idea what's up with the high ports (>10000, can't find any mention in the docs) and those seem like they might be key to getting the deploy to work |
|
bah, on the official images (and consequently, |
|
mm ok seems like progress cause now it's hung on txn confirmation, not the sending txns like it was before...
can't really quite figure out why these high ports still aren't landing in dynamic-port-range though, perhaps that's causing the hang
|
|
trying some additional logging in the client to untangle this, looks like it gets stuck on trying to confirm the transaction, the RPC call to do so seems to never return ... index 759b4a332..bd21fc809 100644
--- a/client/src/rpc_client.rs
+++ b/client/src/rpc_client.rs
@@ -1616,10 +1616,12 @@ impl RpcClient {
signature: &Signature,
commitment_config: CommitmentConfig,
) -> ClientResult<Option<transaction::Result<()>>> {
- let result: Response<Vec<Option<TransactionStatus>>> = self.send(
- RpcRequest::GetSignatureStatuses,
- json!([[signature.to_string()]]),
- )?;
+ println!("get_signature_status_with_commitment");
+ let params = json!([[signature.to_string()]]);
+ println!("params: {:?} ", params);
+ let result: Response<Vec<Option<TransactionStatus>>> =
+ self.send(RpcRequest::GetSignatureStatuses, params)?;
+ println!("result: {:?}", result);
Ok(result.value[0]
.clone()
.filter(|result| result.satisfies_commitment(commitment_config))
|
|
well this is truly bizarre, when I add another log line in the http rpc sender, i actually do get result logged diff --git a/client/src/http_sender.rs b/client/src/http_sender.rs
index 0a7558c6f..8531fe7e9 100644
--- a/client/src/http_sender.rs
+++ b/client/src/http_sender.rs
@@ -118,6 +118,7 @@ impl RpcSender for HttpSender {
let response = {
let client = self.client.clone();
let request_json = request_json.clone();
+ println!("url: ${:?}", self.url);
tokio::task::block_in_place(move || {
client
.post(&self.url)
seems racey |
|
Client works fine on devnet >_> |
|
mm, now I can't get it to run in the container either... |
|
it's like this thing just gives up after a while, truly strange, after fiddling a bit with the test validator image i've been using for debugging, even a simple airdrop transaction won't go through ... edit: OK so |
|
looks like
|
|
mm yea in a totally fresh container the deploy works, makes me think somehow the state gets cooked with a failed deploy |
|
hm, |
|
this is all on ARM Macbook |
nathanleclaire
changed the title
|
Not quite fixed yet unfortunately. def think With latest config I get these results. Hangs on "Waiting for the next block" / "Checking transaction status"
Seeing these warnings in logs. Think they might be relevant, esp. window service port config ... |
|
some notes on this given that it now seems to be fixed (?) on my M1: Solana ports still confuse me. Six UDP ports on there that have me scratching my head. Plus, what I remember from before is that ports higher than 10008, like 10009 were filled too. Still get the warning |
yeah, me too
OK, so using the non-container
solana-test-validatorworks, so yay for containers.The text was updated successfully, but these errors were encountered: