API sometimes unresponsive

Question

Hi

I am running 2 nodes, one of which has mining granted. The non mining node is where I am sending the API requests. Quite often the API becomes unresponsive, as does the CLI, which just uses the API anyway. Eventually it comes back to life. Is there some configuration I still need to do to stop this happening?

Latest version of multichain installed on both nodes.

Thanks

Russell

Answer 1

Thanks for reporting this. Can you please give us some idea of:

a) the load that the blockchain is under (i.e. how many transactions per second, and of what nature).

b) the rate and nature of the API queries you are sending.

c) how often the unresponsive periods last.

d) what you see when it's unresponsive - is it just a failed HTTP connection?

Comments

Hi

a) nothing really, 5 tx an hour!! we are just testing... We are doing issuemorefrom, sendassetfrom and getaddressbalances

b) 30-60 seconds maybe

c) the CLI hangs and eventually produces a response. The api call from our app times out after a minute

Russell

---

Thanks. Could it be a networking issue, or is the CLI (or code making the API requests) on the same host as multichaind?

---

The client is remote, but the clil is on the same box as the master node running multichaind

---

OK, thanks. This does sound like a multichaind problem, but it's not one we've heard of before. Would you be willing to help us see if we can reproduce and debug this? We would need you to stop your node, then zip up the full contents of the blockchain directory (inside ~/.multichain) and email it to us at multichain.debug at gmail dot com. In addition it would be helpful to know the specs of the server on which multichaind is running - operating system and CPU (or if a cloud instance, which). And also to please send the output of these commands, assuming you're running on Linux:

free
df -ah
uptime

Thanks. Hopefully that will give us everything we need to try and understand why this is happening. It's certainly not a common problem.