Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE BLURT!
Sort Order:  

I only have one piece of bad news on that which is that I hit an additional snag with blurt on Pi.

The FC library does not seem to be compatible with arm. I may be able to backport from EOS however.

  ·  4 years ago  ·  

I'm pretty sure you will manage to solve the problems step by step. It should be able to setup a blockchain like Blurt pretty new as well. Without all this old fuzzy Graphene stuff it would be a better option for future.

I installed 64-bit Raspbian OS Beta on my P4-8GB. Is it okay or should I take another OS?

That OS is okay. But in fact I am packaging the chain with the OS so that you don't need to install and configure because that's a pain.

  ·  4 years ago  ·   (edited)

Where do I get this image? I guess it is Megadrive with SOS, ready to be burned on a SD.
Can't find it, I only find SOS.

You're right.

Let me see if my work from last night is finished cooking

  ·  4 years ago  ·   (edited)

You remember the tree commands, you wrote to me in an answer before?

docker run --net=host faddat/blurt-pi /usr/bin/blurt --data-dir /blurtd

This setup doesn't work on Raspian 64 bit. The container doesn't start. only "dockerd" was running without starting a container. A kind of download with 27 packages where pulled but no start of chain followed.

After all I tried to start the node by a command line, I saw in one of you movies:

docker run --net=host faddat/blurt-pi /usr/bin/blurt --data-dir

It's nearly the same instruction, but without the "/blurtd" at the end. It worked! Now I saw the container's start and then the Blockchain came to download and started to complete itself, block by block, very fast, line by line until hardfork 1. Then it continued but very, very, very slow. The pattern of screen messages changed totally from one line per block to messages like the following pattern:

{"new_block":{"previous":"000452ff33605382d2caf1c8a67261ccc3af4edd","timestamp":"2020-07-14T02:24:57","witness":"empato365","transaction_merkle_root":"0000000000000000000000000000000000000000","extensions":[],"witness_signature":"20c60fbb7458f2c6561b2e44ddd9c87a3088b8670a141e9672dbd297f7283f6e344b0f3cd50071b1ecad192ccf1deedd84fa7eea1b1a2bcc187d336951e16bc71b","transactions":[]}}   
database.cpp:759 operator()
3411634ms p2p_plugin.cpp:229            handle_block         ] Error when pushing block:
10 assert_exception: Assert Exception
item->num > std::max<int64_t>( 0, int64_t(_head->num) - (_max_size) ): attempting to push a block that is too old
    {"item->num":284433,"head":284790,"max_size":20}
    fork_database.cpp:57 _push_block

    {}
    database.cpp:871 _push_block

Six hours later, the downloads are on timestamp September 07 "already", and the pattern of messages changed again. Now the blocks are shown with any transaction inside, and still such slow as after hardfork 1.

What do you recommend? Should I let it go on? My opinion is to stop it and wait for your Megadrive image. But I have no experience with building nodes. Maybe this described behaviour is quite normal and it takes two or three days to complete the node. What do you recommend?

  ·  4 years ago  ·   (edited)

Okay I'll tell you about my experience, and thanks so much for testing.

So I updated the image and I think that the correct folder is now no longer /Blurtd

It's /blurt

You may wish to try with --data-dir /blurt

Please feel free to post any and all logs.

I can make an image for you In fact I will, that runs without the operating system. Right now I'm working on all the services that need to start when the raspberry Pi boots.

I really appreciate your testing.

Okay also about sync times, on my machines at hetzner.de I sync in "a few hours". On my AWS graviton instance, I sync in 6 hours.

My pi took 24 hours to sync.

I also just realized that there may be a way better way to ship these images, much more efficient. Let me see what tonight's session brings.

  ·  4 years ago  ·  

Thanks for your attention. Since two hrs the sync is running a little bit faster and the one line messages has come back. These big message blocks with the pushes come only sometimes now. So there are less stops at the moment. I have catched approx the half of the chain now (now around Block Num. 2300000 of approx 4510000).

Ok, I let it run. I've read a lot of times, people can't believe how much time it takes, so I think the absence of speed is quite normal.

Am I able to use this download for my later trials too, or do I have to download the chain during every new experiment?