Are you running electrs as a background process or are you running it in a terminal window and can see all of it's debug output?
When you run Electrum, do you see any notifications in the electrs output that shows any connection attempts? You might also want to run manually running Electrum from a terminal with the 'debug' option (-v) to see why it isn't able to connection.
There is also an option in Electrum to log to disk: Tools menu > Preferences > General tab, and tick "Write logs to file"
Topic: An efficient re-implementation of Electrum Server in Rust (Read 627 times)
Unable to get Electrum wallet to connect to electrs
Unsure if this is the right thread to ask this, however I couldn't find a thread that related to this issue so here goes -
Have electrs running (confirmed with: sudo systemctl status electrs). Have installed Electrum wallet, & edited the electrum.desktop file as follows to connect to localhost -
Network Server tab in Electrum wallet shows "Server: localhost 50001" & all other servers greyed out, however within Network Overview, Status: shows "Not connected" & Server: shows "None"
Perhaps there are config edits I need to make elsewhere for Electrum to connect to electrs?
Any suggestions greatly appreciated! Thank you in advance.
Unsure if this is the right thread to ask this, however I couldn't find a thread that related to this issue so here goes -
Have electrs running (confirmed with: sudo systemctl status electrs). Have installed Electrum wallet, & edited the electrum.desktop file as follows to connect to localhost -
Code:
Exec=sh -c "PATH=\"\\$HOME/.local/bin:\\$PATH\"; electrum --oneserver --server localhost:50001:t %u"
Network Server tab in Electrum wallet shows "Server: localhost 50001" & all other servers greyed out, however within Network Overview, Status: shows "Not connected" & Server: shows "None"
Perhaps there are config edits I need to make elsewhere for Electrum to connect to electrs?
Any suggestions greatly appreciated! Thank you in advance.
Quote
So, try this:
Code:
cargo run --release -- -vvv --timestamp --db-dir ./db --daemon-dir /mnt/HDD/ --electrum-rpc-addr="127.0.0.1:50001" --cookie-file /home/user/.bitcoin/.cookie
It worked! Electrs is up & running, financial self-sovereignty is within very close reach. Yes my apologies, I meant to refer to electrs not Electrum in my previous comment - rookie mistake.
As a non-techie noob I'm blown away by this world of open source development & the remarkable minds working together to make it all happen.
HCP Sir, a MASSIVE thank you for walking me through this. Means a lot. Hope to be able to contribute in some way myself in future.
Great, thank you for suggesting! Doesn't seem to do the trick unfortunately. I assume that --db-dir ./db within the argument refers to the Electrum database, so I've left this unchanged given I have Electrs located in the default /home/electrs/ location.
Did you mean Electrum database... or electrs database? 
db-dir is the location where electrs stores it's database.Quote
The .cookie file is not produced within the /mnt/HDD/ folder but instead within /Home/User/.bitcoin, together with the bitcoin.conf file, & I have the following argument within the bitcoin.conf file that sets /mnt/HDD/ as the data directory: "datadir=/mnt/HDD"
I have the following argument - "rpccookiefile=/home/user/.bitcoin/.cookie" - within the bitcoin.conf file to create the .cookie file there so that Bitcoind will run from the default /Home/User/.bitcoin directory but use the /mnt/HDD to store the blocks.
Ahhh ok, yeah... you've added an extra layer of complexity by moving some of the Bitcoin data to your /mnt/HDD location, but then leaving some of it in the default "/home/user/.bitcoin" location.I have the following argument - "rpccookiefile=/home/user/.bitcoin/.cookie" - within the bitcoin.conf file to create the .cookie file there so that Bitcoind will run from the default /Home/User/.bitcoin directory but use the /mnt/HDD to store the blocks.
By default, electrs simply looks for the cookie in the daemon-dir... If you have the cookie in a custom location, you'd need to specify that. There is a cookie-file argument that you can use for this.
So, try this:
Code:
cargo run --release -- -vvv --timestamp --db-dir ./db --daemon-dir /mnt/HDD/ --electrum-rpc-addr="127.0.0.1:50001" --cookie-file /home/user/.bitcoin/.cookie
Ok... so I use a shell script to start it up... it's basically one line:
As you can see... I've got --daemon-dir added into the "cargo run" command... and specified the path to my Bitcoin blocks folder... "/mnt/e/Bitcoin" (I've also put the electrs db-dir in a custom location too)
Code:
cargo run --release -- -vvv --timestamp --db-dir /mnt/e/electrs --daemon-dir /mnt/e/Bitcoin --electrum-rpc-addr="127.0.0.1:50001"
As you can see... I've got --daemon-dir added into the "cargo run" command... and specified the path to my Bitcoin blocks folder... "/mnt/e/Bitcoin" (I've also put the electrs db-dir in a custom location too)
Great, thank you for suggesting! Doesn't seem to do the trick unfortunately. I assume that --db-dir ./db within the argument refers to the Electrum database, so I've left this unchanged given I have Electrs located in the default /home/electrs/ location. Thus my code is as follows -
Code:
cargo run --release -- -vvv --timestamp --db-dir ./db --daemon-dir /mnt/HDD/ --electrum-rpc-addr="127.0.0.1:50001"
This returns the following recurring error:
"WARN - reconnecting to bitcoind: failed to read cookie from /mnt/HDD/.cookie"
.
The .cookie file is not produced within the /mnt/HDD/ folder but instead within /Home/User/.bitcoin, together with the bitcoin.conf file, & I have the following argument within the bitcoin.conf file that sets /mnt/HDD/ as the data directory: "datadir=/mnt/HDD"
I have the following argument - "rpccookiefile=/home/user/.bitcoin/.cookie" - within the bitcoin.conf file to create the .cookie file there so that Bitcoind will run from the default /Home/User/.bitcoin directory but use the /mnt/HDD to store the blocks.
I've tried setting /Home/User/.bitcoin as the --daemon-dir in your suggested argument however this returned the original error, where Electrs appears to look for blocks within /Home/User/.bitcoin instead of /mnt/HDD.
Hope this makes sense, very new to Ubuntu/Linux & setting up a node etc. How to proceed from here?
Thank you in advance!
Ok... so I use a shell script to start it up... it's basically one line:
As you can see... I've got --daemon-dir added into the "cargo run" command... and specified the path to my Bitcoin blocks folder... "/mnt/e/Bitcoin" (I've also put the electrs db-dir in a custom location too)
Code:
cargo run --release -- -vvv --timestamp --db-dir /mnt/e/electrs --daemon-dir /mnt/e/Bitcoin --electrum-rpc-addr="127.0.0.1:50001"
As you can see... I've got --daemon-dir added into the "cargo run" command... and specified the path to my Bitcoin blocks folder... "/mnt/e/Bitcoin" (I've also put the electrs db-dir in a custom location too)
"Before anyone asks... you use --daemon-dir to specify your custom Bitcoin directory location 
"
Where do you specify this exactly? I'm an Ubuntu / Linux noob trying to set up Electrum Rust using laptop + external HDD, & have run into a problem I can't seem to solve -
Node is synced to the tip & fully indexed, with blocks & data stored on external HDD, while bitcoin.conf & .cookie files appear in .bitcoin folder in Home directory. However in my first Electrum Rust "usage" ie. index sync ($ cargo run --release -- -vvv --timestamp --db-dir ./db --electrum-rpc-addr="127.0.0.1:50001"), as per the "Usage" section (https://github.com/romanz/electrs/blob/master/doc/usage.md) I get the following error -
"2020-03-02T16:09:42.188+10:30 - INFO - indexing 0 blk*.dat files
2020-03-02T16:09:42.189+10:30 - DEBUG - found 0 indexed blocks
2020-03-02T16:10:03.639+10:30 - DEBUG - applying 619796 new headers from height 0
thread 'bulk_writer' panicked at 'no indexed header found', src/libcore/option.rs:1188:5
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace.
thread 'main' panicked at 'writer panicked: Any', src/libcore/result.rs:1188:5"
My data directory is on my external HDD, however it appears Electrs is looking for blocks in the default directory?
Any help would be very much appreciated!
"Where do you specify this exactly? I'm an Ubuntu / Linux noob trying to set up Electrum Rust using laptop + external HDD, & have run into a problem I can't seem to solve -
Node is synced to the tip & fully indexed, with blocks & data stored on external HDD, while bitcoin.conf & .cookie files appear in .bitcoin folder in Home directory. However in my first Electrum Rust "usage" ie. index sync ($ cargo run --release -- -vvv --timestamp --db-dir ./db --electrum-rpc-addr="127.0.0.1:50001"), as per the "Usage" section (https://github.com/romanz/electrs/blob/master/doc/usage.md) I get the following error -
"2020-03-02T16:09:42.188+10:30 - INFO - indexing 0 blk*.dat files
2020-03-02T16:09:42.189+10:30 - DEBUG - found 0 indexed blocks
2020-03-02T16:10:03.639+10:30 - DEBUG - applying 619796 new headers from height 0
thread 'bulk_writer' panicked at 'no indexed header found', src/libcore/option.rs:1188:5
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace.
thread 'main' panicked at 'writer panicked: Any', src/libcore/result.rs:1188:5"
My data directory is on my external HDD, however it appears Electrs is looking for blocks in the default directory?
Any help would be very much appreciated!
Thank you, the update went smoothly. The Cargo build only took 45 seconds or so.
The underlying storage format didn't change, so the instructions above should work 
What would you recommend are the steps to update to the newest release v0.7 (in my case I have v0.62 from early June) ?
I have not made any local changes to the code.
I have not made any local changes to the code.
Code:
$ git clone https://github.com/romanz/electrs
$ cd electrs
$ git fetch origin
$ git status
$ git pull
$ cargo build --release
It should work with latest Bitcoin Core (0.16+).
Didn't test with Parity or with Bitcoin Cash nodes...
Didn't test with Parity or with Bitcoin Cash nodes...
Which full node implementations of bitcoin and bitcoin* cash are compatible with this Electrum server? I tried Parity, so that I could run bitcoin and bitcoin cash while only having to upgrade one codebase. However, ElectrumX will not speak to Parity. From my perspective, Parity and bitcoin core follow the same network protocol, but within my localhost they use a different database formats. Hence why ElectrumX rejects Parity's RPC responses with:
Jan 30 05:55:12 electrumx_server[4937]: ERROR:Daemon:Supplied content type is not allowed. Content-Type: application/json is required
I didn't see any specs for the full node implementation compatibility on either https://github.com/romanz/electrs/blob/master/doc/usage.md or https://github.com/kyuupichan/electrumx/tree/master/docs . Also, is there a way to run one full node implementation Thanks for reading.
*I use lower case to denote that I am referring to the bitcoin/bitcoin cash networks and not the units of account.
Code:
Jan 30 05:55:12 electrumx_server[4937]: ERROR:root:task crashed:Jan 30 05:55:12 electrumx_server[4937]: Traceback (most recent call last):
Jan 30 05:55:12 electrumx_server[4937]: File "/home/electrum/.local/lib/python3.7/site-packages/aiorpcx/util.py", line 118, in
.
.
I didn't see any specs for the full node implementation compatibility on either https://github.com/romanz/electrs/blob/master/doc/usage.md or https://github.com/kyuupichan/electrumx/tree/master/docs . Also, is there a way to run one full node implementation Thanks for reading.
*I use lower case to denote that I am referring to the bitcoin/bitcoin cash networks and not the units of account.
Thanks!
This project was designed for personal use (e.g. optimized for HDD, no DoS prevention mechanisms).
Currently, it doesn't utilize SSD for better performance (so it would be hard to support thousands users), but there is a WIP branch for a new (more SSD-friendly) indexing schema here:
https://github.com/romanz/electrs/commits/new-index
This project was designed for personal use (e.g. optimized for HDD, no DoS prevention mechanisms).
Currently, it doesn't utilize SSD for better performance (so it would be hard to support thousands users), but there is a WIP branch for a new (more SSD-friendly) indexing schema here:
https://github.com/romanz/electrs/commits/new-index
Interesting project! Electrum server could certainly be improved on.
Is this software for running a personal server or could it be used to run a public server?
How well would it handle the load of thousands simultaneous users, asking for many addresses each?
Is this software for running a personal server or could it be used to run a public server?
How well would it handle the load of thousands simultaneous users, asking for many addresses each?
Sorry for the delayed response - you can use a TLS-terminating proxy (such as https://hitch-tls.org/).
See https://github.com/romanz/electrs/commit/288cda518c3df6d49ed6d27fcd8f9f10d69a84cd#diff-9d70860c70981206f00c69a429b69ad1.
See https://github.com/romanz/electrs/commit/288cda518c3df6d49ed6d27fcd8f9f10d69a84cd#diff-9d70860c70981206f00c69a429b69ad1.
I guess the lack of SSL isn't a "big deal", since I'm connecting to localhost... but are you planning on implementing SSL in case I'd like to host it on a separate/remote box?
Cool, glad to hear that it works
Has been slow going... My router died the other day and after a day of being bounced around various departments at my ISP to get a replacement, a friend loaned me one and I managed to get back online...
Since then the various Ubuntu and Rust repositories have been spitting out random errors... So I think my connection still isn't quite right.
I think I finally managed to get it all synced up and set it compiling before heading to work... Will check it after my shift
EDIT: ok, so I managed to get it compiled and running and it is picking up new blocks:
However, when I attempt to connect with Electrum (v3.2.2)... it starts generating "UTF8" errors
Any ideas? NOTE: also created an issue on GitHub
EDIT2: as per the note on Github... it is related to SSL connections... currently there is no way to disable SSL in the GUI... but starting Electrum and forcing to use TCP with the following command:
And it all connects and syncs up... pretty happy to have my own Electrum server now
Since then the various Ubuntu and Rust repositories have been spitting out random errors... So I think my connection still isn't quite right.
I think I finally managed to get it all synced up and set it compiling before heading to work... Will check it after my shift
EDIT: ok, so I managed to get it compiled and running and it is picking up new blocks:
Code:
2018-07-20T19:36:14.665+12:00 - DEBUG - Server listening on 127.0.0.1:42024
2018-07-20T19:36:14.669+12:00 - DEBUG - Running accept thread
2018-07-20T19:36:14.682+12:00 - DEBUG - BlockchainInfo { chain: "main", blocks: 532757, headers: 532757, bestblockhash: "00000000000000000021489dbeba2486cb06231787a01248d5279b478d44d29d", size_on_disk: 199680846175, pruned: false }
2018-07-20T19:36:14.684+12:00 - DEBUG - NetworkInfo { version: 160100, subversion: "/Satoshi:0.16.1/" }
2018-07-20T19:36:14.684+12:00 - DEBUG - opening DB at "/mnt/s/db/mainnet"
2018-07-20T19:36:15.476+12:00 - INFO - listing block files at "/mnt/e/Bitcoin/blocks/blk*.dat"
2018-07-20T19:36:15.504+12:00 - INFO - indexing 1321 blk*.dat files
2018-07-20T19:37:08.956+12:00 - DEBUG - applying 532758 new headers from height 0
2018-07-20T20:07:53.198+12:00 - DEBUG - skipping block 000000000000000001b982e5989a72cd799130f71cf3acd6db6214de024a8ad2
2018-07-20T20:07:53.669+12:00 - DEBUG - skipping block 000000000000000000352b9ad3494acf88b0aa73a4c6f8987bab247e6c32d9ca
2018-07-20T20:08:13.935+12:00 - DEBUG - skipping block 0000000000000000009085be97279a7ab4d20e983be9cb20fe0c2e0d49313cdd
...
snip Bunch of "DEBUG - skipping block lines"
...
2018-07-20T20:28:43.430+12:00 - DEBUG - no more blocks to index
2018-07-20T20:28:44.543+12:00 - DEBUG - last indexed block: best=00000000000000000021489dbeba2486cb06231787a01248d5279b478d44d29d height=532757 @ 2018-07-20T07:32:29Z
2018-07-20T20:28:44.774+12:00 - INFO - starting full compaction
2018-07-20T21:38:04.565+12:00 - INFO - finished full compaction
2018-07-20T21:38:09.008+12:00 - DEBUG - applying 532758 new headers from height 0
2018-07-20T21:38:09.518+12:00 - DEBUG - downloading new block headers (532758 already indexed) from 0000000000000000000dc8a8b42d286c1796337a3cc44622143320e98a40415c
2018-07-20T21:38:09.539+12:00 - INFO - best=0000000000000000000dc8a8b42d286c1796337a3cc44622143320e98a40415c height=532779 @ 2018-07-20T09:26:49Z (22 left to index)
2018-07-20T21:38:13.016+12:00 - DEBUG - applying 22 new headers from height 532758
2018-07-20T21:38:13.822+12:00 - INFO - RPC server running on 127.0.0.1:50001
2018-07-20T21:55:05.336+12:00 - DEBUG - downloading new block headers (532780 already indexed) from 00000000000000000003972c3d95015978a4b7fa5662aee1417b677058178318
2018-07-20T21:55:05.338+12:00 - INFO - best=00000000000000000003972c3d95015978a4b7fa5662aee1417b677058178318 height=532780 @ 2018-07-20T09:54:40Z (1 left to index)
2018-07-20T21:55:06.067+12:00 - DEBUG - applying 1 new headers from height 532780
However, when I attempt to connect with Electrum (v3.2.2)... it starts generating "UTF8" errors
Code:
2018-07-20T23:40:28.775+12:00 - INFO - [127.0.0.1:61899] connected peer
2018-07-20T23:40:28.783+12:00 - INFO - [127.0.0.1:61899] shutting down connection
2018-07-20T23:40:28.783+12:00 - ERROR - [127.0.0.1:61899] receiver failed: invalid UTF8 "\u{16}\u{3}\u{1}\u{2}\u{0}\u{1}\u{0}[b][REDACTED][/b]\n": FromUtf8Error { bytes: [22, 3, 1, 2, 0, 1, 0, [b][REDACTED][/b]], error: Utf8Error { valid_up_to: 8, error_len: Some(1) } }
2018-07-20T23:40:28.786+12:00 - INFO - [127.0.0.1:61900] connected peer
2018-07-20T23:40:28.786+12:00 - INFO - [127.0.0.1:61899] disconnected peer
2018-07-20T23:40:28.792+12:00 - INFO - [127.0.0.1:61900] shutting down connection
2018-07-20T23:40:28.793+12:00 - ERROR - [127.0.0.1:61900] receiver failed: invalid UTF8 "\u{16}\u{3}\u{1}\u{2}\u{0}\u{1}\u{0}[b][REDACTED][/b]\n": FromUtf8Error { bytes: [22, 3, 1, 2, 0, 1, 0, [b][REDACTED][/b]], error: Utf8Error { valid_up_to: 8, error_len: Some(1) } }
2018-07-20T23:40:28.794+12:00 - INFO - [127.0.0.1:61900] disconnected peer
Any ideas? NOTE: also created an issue on GitHub
EDIT2: as per the note on Github... it is related to SSL connections... currently there is no way to disable SSL in the GUI... but starting Electrum and forcing to use TCP with the following command:
Code:
"C:\Program Files (x86)\Electrum\electrum-3.2.2.exe" --server=127.0.0.1:50001:t
And it all connects and syncs up... pretty happy to have my own Electrum server now
Quote
I'll update how I get on.
Thanks a lot!
Please let me know if there are any issues/problems
Before anyone asks... you use --daemon-dir to specify your custom Bitcoin directory location
All commandline args are listed here: https://github.com/romanz/electrs/blob/master/src/config.rs
I'm currently attempting to see if this will work on Ubuntu 16.04 running on the Windows Linux Subsystem... combined with Bitcoin Core running under Windows. I'll update how I get on.
All commandline args are listed here: https://github.com/romanz/electrs/blob/master/src/config.rs
I'm currently attempting to see if this will work on Ubuntu 16.04 running on the Windows Linux Subsystem... combined with Bitcoin Core running under Windows. I'll update how I get on.
Jump to: