This repository is a mirror of https://git.openwrt.org/openwrt/openwrt.git It is for reference only and is not active for check-ins. We will continue to accept Pull Requests here. They will be merged via staging trees then into openwrt.git.
Screenshot, because error does not seem to occur in syslog; on physical screen only (if emulated screen such as qemu counts).
disclaimer:
I readed procd sources partially, all the information about procd code might not be accurate.
Failed to connect to ubus. By using more verbose boot logging, it is revealed that procd is attempting to execute /etc/init.d/ubus running when this error occurs. This file does not exist, because ubusd does not need initscript, procd starts it. I created a quite empty init script for testing purposes, which doesn't start anything. But has running function, yet error existed, as after that succeeding it procd attempts to set details of started service of ubus, to ubus. Problem is that ubus daemon is not yet ready/does not accept this connection, hence the error.
Yet, when I use ubus call service list - proper details about ubus service do exist, I didn't search details where these are set, as clearly when described error occurs, these won't be set - maybe ubus sets it's own state independently, or details are updated later? Here's output of ubus service details:
I assume ubus startup is made in state.c of procd near line 155 - and fix to this would be to alter ubus startup to ignore updating ubus details as a service on the "first run".. Only when updated etc..
I know, it's merely a cosmetic bug. But yet, it's a bug...
OpenWrt version
r24069-954142f477
OpenWrt target/subtarget
x86/64
Device
Red Hat KVM
Image kind
Self-built image
Steps to reproduce
boot without log suppression to quiet level
Actual behaviour
unnecessary error
Expected behaviour
success without error and not to attempt update details of ubus, when ubus is not ready/available yet
Additional info
tagging here few people who've been lately involved with project, as original maintainer doesn't seem to be available at GitHub for the moment.
@nbd168 @dangowrt @hauke
Diffconfig
unrelevant
Terms
[X] I am reporting an issue for OpenWrt, not an unsupported fork.
Describe the bug
Screenshot, because error does not seem to occur in syslog; on physical screen only (if emulated screen such as qemu counts).
disclaimer: I readed procd sources partially, all the information about procd code might not be accurate.
Failed to connect to ubus. By using more verbose boot logging, it is revealed that procd is attempting to execute
/etc/init.d/ubus running
when this error occurs. This file does not exist, because ubusd does not need initscript, procd starts it. I created a quite empty init script for testing purposes, which doesn't start anything. But has running function, yet error existed, as after that succeeding it procd attempts to set details of started service of ubus, to ubus. Problem is that ubus daemon is not yet ready/does not accept this connection, hence the error.Yet, when I use
ubus call service list
- proper details about ubus service do exist, I didn't search details where these are set, as clearly when described error occurs, these won't be set - maybe ubus sets it's own state independently, or details are updated later? Here's output of ubus service details:Details seem to be correct.
I assume ubus startup is made in state.c of procd near line 155 - and fix to this would be to alter ubus startup to ignore updating ubus details as a service on the "first run".. Only when updated etc..
I know, it's merely a cosmetic bug. But yet, it's a bug...
OpenWrt version
r24069-954142f477
OpenWrt target/subtarget
x86/64
Device
Red Hat KVM
Image kind
Self-built image
Steps to reproduce
boot without log suppression to quiet level
Actual behaviour
unnecessary error
Expected behaviour
success without error and not to attempt update details of ubus, when ubus is not ready/available yet
Additional info
tagging here few people who've been lately involved with project, as original maintainer doesn't seem to be available at GitHub for the moment.
@nbd168 @dangowrt @hauke
Diffconfig
Terms