Hello.
I think you need to use some of this configuration properties:
https://docs.symbolplatform.com/guides/network/configuring-node-properties.html#node-properties
I see there is maxUnlockedAccounts where you can setup the max number of accounts that can be delegated to your node. I do not see any whitelist or blacklist options in the list of properties.
Thanks freigeist for the hint. Another question (I’m still finding a reason, why symbol-bootstrap is failing on some HW / in some VM). On one of failing servers (yeah, without AVX/AVX2 instructions, it would be useful, if somebody tells to us, if these instructions are necessary or not) I observe this in the log ::
rest-gateway | verbose: connecting to node:7900 {“timestamp”:“2021-04-20T14:56:19.264Z”}
rest-gateway | verbose: error raised by node:7900 connection: getaddrinfo ENOTFOUND node {“errno”:“ENOTFOUND”,“code”:“ENOTFOUND”,“syscall”:“getaddrinfo”,“hostname”:“node”,“stack”:“Error: getaddrinfo ENOTFOUND node\n at GetAddrInfoReqWrap.onlookup [as oncomplete] (dns.js:66:26)”,“timestamp”:“2021-04-20T14:56:19.265Z”}
rest-gateway | verbose: error raised by node:7900 connection: getaddrinfo ENOTFOUND node {“errno”:“ENOTFOUND”,“code”:“ENOTFOUND”,“syscall”:“getaddrinfo”,“hostname”:“node”,“stack”:“Error: getaddrinfo ENOTFOUND node\n at GetAddrInfoReqWrap.onlookup [as oncomplete] (dns.js:66:26)”,“timestamp”:“2021-04-20T14:56:19.266Z”}
rest-gateway | error: caught error 503 connection failed {“jse_shortmsg”:“connection failed”,“jse_info”:{},“body”:{“code”:“ServiceUnavailable”,“message”:“connection failed”},“stack”:“ServiceUnavailableError: connection failed\n at Object.createServiceUnavailableError (/app/catapult-rest/rest/_build/server/errors.js:56:44)\n at TLSSocket. (/app/catapult-rest/rest/_build/connection/connectionService.js:104:20)\n at TLSSocket.emit (events.js:327:22)\n at net.js:674:12\n at TCP.done (_tls_wrap.js:567:7)”,“timestamp”:“2021-04-20T14:56:19.266Z”}
It’s strange, as this machine has /etc/hosts present (I’ve been found in some forums, that missing localhost definition could cause this issue), resolving DNS requests, has normal/full access to the Internet, etc…
Now I’m just tried this on processor with AVX (but not AVX2) instructions. Died the same container as Dimas observed (symbolplatform/symbol-server:gcc-1.0.0.0), in the log I can observe ::
RUNNING server NORMAL node
!!! Starting server…
/symbol-commands/start.sh: line 126: 14 Illegal instruction (core dumped) $catapultAppFolder/bin/$processName “$config”
I think that lot of us will highly appreciate, if some of devs will explain these crashes, how it is (or isn’t) related to AVX/AVX2 instructions, etc. If we want to build a large Symbol network, answers to our feedback will be really great (and important) …
This issue is more than month without clear answers, which is really sad. So, again, could be some of devs (are the devs reading this forum? Interesting question…) so kind and explain ::
Is the symbol-bootstrap really related on AVX2 (or some another processor instruction set) ?
If it is, why’s that ?
If it is, is there or could be there some workaround, how to run symbol-bootstrap at least on the processors with “just” AVX (lot of powerful Intel-based servers, with computing capacity fair enough for running Symbol node, is “AVX only”) ?
Thanks in advance.
hi,
I have the same problem after move my machine from one server to another… I can reinstall centos 8 and symbol-bootstrap but
how can I save actual node config? (config ok but addresses keys etc)
another question, is it better to install Centos 8 (eol end 2021) or Centos 8 Stream for new installation?
thanks
