1

I'm trying to run bitcoind on an AWS debian instance. Running with the -daemon option would terminate with no indication as to why so I decided to run it under a screen. I started it last night (without options) and disconnected from the screen and verified that it was working. This morning the server had stopped so I reconnected to the screen session it was running on and the only message bitcoind gave was "Killed".

There's nothing in the logs that seems to indicate how or why this is happening and I'm at a bit of a loss. Can anyone give me some recommendations or insight on how to determine the cause of the issue?

  • 1
    This often occurs when the machine runs out of memory. If so, there would likely be a message in one of your system logs (not the bitcoind log). How much memory does your instance have? – Nate Eldredge Jan 30 '14 at 18:07
  • 1
    @SunWKim: I don't know exactly how AWS works, but it seems unlikely that this is due to Amazon's interference. Providers typically don't mess with what's running on a virtual machine; you've paid for the instance and you're entitled to use all its resources. If there were a problem, I'd expect them to shut down the instance completely, not go around killing processes on it. – Nate Eldredge Jan 30 '14 at 18:10
  • @NateEldredge - Thanks for the memory suggestion. I have ~589 mb and linux is reporting about 6mb free when the server is running so you're probably correct. You mention there is a system log that would indicate an out of memory exception? I know it's most likely the issue but for my own curiosity where would I locate it? (Linux newbie) – Spencer Ruport Jan 30 '14 at 19:21
  • Is this a free usage tier or AWS you have paid for? From my understanding, if you have to pay for AWS you are losing money since you will not mine enough in the long run to pay for the computing cost. – Sun Jan 30 '14 at 19:57
  • Yeah, that's probably too little memory. For reference, bitcoin-qt is using about 400MB on my machine right now, and of course you need some for the rest of the system. For logs, I'm not 100% sure about Debian but I would look in /var/log/syslog, or if you haven't rebooted since the incident, run dmesg |less. – Nate Eldredge Jan 30 '14 at 20:36
3

Configure some swap space, you are probably running out of memory, check with dmesg on your instance for additional messages.

0

If you are downloading the blockchain you will likely run out of hard drive space. Most default instances start with only 8 GB of storage. SSH into the box and check disk utilization.

df -h
  • Good point but in my case I have attached a 30 gig volume and it's about half full. – Spencer Ruport Jan 31 '14 at 2:21
  • 2
    Also, running out of disk space wouldn't likely result in the process getting SIGKILL. It would typically fail more gracefully with an I/O error. – Nate Eldredge Jan 31 '14 at 2:56
  • no space left on device. – Giovanni Toraldo Feb 1 '14 at 15:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.