What's new

Entware Entware

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

Dear Voxel,
After upgrade unbound to the most recent version (1.13.0) there are issues with accessing domain 'internic.net'.
unbound-host -C /opt/etc/unbound/unbound.conf -v -d internic.net. -t SOA
internic.net. has SOA record sns.dns.icann.org. noc.dns.icann.org. 2020121130 3600 1800 604800 3600 (BOGUS (security failure))
validation failure <internic.net. SOA IN>: No DNSKEY record for key internic.net. while building chain of trust

There are many "Operation not supported" in unbound log before getting above error 'No DNSKEY record'.
Jan 17 22:51:49 libunbound[18715:0] error: tcp sendmsg: Operation not supported for 199.4.138.53 port 53
Jan 17 22:51:49 libunbound[18715:0] error: tcp sendmsg: Operation not supported for 192.82.133.30 port 53
Jan 17 22:51:49 libunbound[18715:0] error: tcp sendmsg: Operation not supported for 192.82.133.30 port 53
Jan 17 22:51:49 libunbound[18715:0] error: tcp sendmsg: Operation not supported for 192.42.178.30 port 53
Jan 17 22:51:49 libunbound[18715:0] error: tcp sendmsg: Operation not supported for 192.42.178.30 port 53
Jan 17 22:51:49 libunbound[18715:0] info: Capsforid: timeouts, starting fallback
Jan 17 22:51:49 libunbound[18715:0] error: tcp sendmsg: Operation not supported for 192.42.177.30 port 53
Jan 17 22:51:49 libunbound[18715:0] error: tcp sendmsg: Operation not supported for 192.42.177.30 port 53
Jan 17 22:51:49 libunbound[18715:0] error: tcp sendmsg: Operation not supported for 202.12.31.53 port 53

Is it possible to get access to previous version(1.11.0) while investigating?
Do you know what problem could be?

There is new feature: "upstream TCP and TLS query reuse".
NLnet Labs - Unbound - Download

Thanks,
 
Is it possible to get access to previous version(1.11.0) while investigating?
It is possible. Usually I keep few previous versions:

1610970815913.png


So it is possible e.g. to manually download unbound*.ipk and to install them from IPK-file by Entware's opkg. Or to change /opt/etc/opkg.conf to switch completely to previous version of Entware.


Do you know what problem could be?
Sorry, I don't. I do not use unbound directly.

Good luck,
Voxel.
 
Hi Voxel,

Is it possible for you to add "php7-pecl-imagick" package?
I needed it for wordpress.
I've passed your request to Entware team. I think it will be possible because "php7-pecl-imagick" is included into OpenWRT packages now:


Voxel.
 
It is possible. Usually I keep few previous versions:

View attachment 29634

So it is possible e.g. to manually download unbound*.ipk and to install them from IPK-file by Entware's opkg. Or to change /opt/etc/opkg.conf to switch completely to previous version of Entware.



Sorry, I don't. I do not use unbound directly.

Good luck,
Voxel.
Thanks Voxel,
I have been able to fix the problem with the unbound 1.13.
It was parameter 'edns-buffer-size' - become default at value '1232' in version 1.12. It was not present in my unbound configuration. At the moment put it to 4096.
Thanks again!
 
Info for Entware users: my version of Entware is upgraded.

It is enough to run:

Code:
/opt/bin/opkg update
/opt/bin/opkg upgrade

to upgrade your version.

Voxel.
Thanks @Voxel. The update/upgrade went smoothly.

It was also great that it detected the nginx config file that I was experimenting with, and did not overwrite it:

Code:
Collected errors:
 * resolve_conffiles: Existing conffile /opt/etc/nginx/nginx.conf is different from the conffile in the new package. The new conffile will be placed at /opt/etc/nginx/nginx.conf-opkg.
 
I will revert to previous version.

It appears that htop does not work properly in latest release (shows no processes), and busybox 1.33.0-1 traceroute is buggy (as reported here: http://forum.tinycorelinux.net/index.php?topic=24849.0 )
Looks like Entware team did not fully check this release...

PS: thanks @Voxel to leave access to previous releases. I downgraded only htop and busybox; works fine.
 
Last edited:
Looks like Entware team did not fully check this release...

It is practically impossible for Entware team (3-4 guys as far as I know) to test everything. Currently Entware consists of 2654 packages... Only compilation (4 parallel threads) takes for me about 6-7 hours... Compare with my firmware for R7800: 377 packages compiled during one hour, single thread.

As a rule sources for Entware is OpenWRT GPL. OpenWRT has (well, as far as I know) relatively big amount of developers responsible for such or such project. E.g. busybox: MAINTAINER:=Felix Fietkau <nbd@nbd.name>. htop: MAINTAINER:=Etienne CHAMPETIER <champetier.etienne@gmail.com>. Entware guys could report the issue to OpenWRT but should sync their packages with OpenWRT.

It is me who makes such brave experiments with packages/compilers/tools in my firmware ;-) because I am able to test almost everything. Entware is used in a lot of various gadgets (mips, armv5, armv6, armv7 with kernel 2.6, armv7 with kernet 3.x, x86 etc.) A lot of thanks to these guys.

Voxel.
 
It is practically impossible for Entware team (3-4 guys as far as I know) to test everything. Currently Entware consists of 2654 packages... Only compilation (4 parallel threads) takes for me about 6-7 hours... Compare with my firmware for R7800: 377 packages compiled during one hour, single thread.

As a rule sources for Entware is OpenWRT GPL. OpenWRT has (well, as far as I know) relatively big amount of developers responsible for such or such project. E.g. busybox: MAINTAINER:=Felix Fietkau <nbd@nbd.name>. htop: MAINTAINER:=Etienne CHAMPETIER <champetier.etienne@gmail.com>. Entware guys could report the issue to OpenWRT but should sync their packages with OpenWRT.

It is me who makes such brave experiments with packages/compilers/tools in my firmware ;-) because I am able to test almost everything. Entware is used in a lot of various gadgets (mips, armv5, armv6, armv7 with kernel 2.6, armv7 with kernet 3.x, x86 etc.) A lot of thanks to these guys.

Voxel.
I get it :)

It seems that the busybox traceroute issue was already reported, but I will double check.
I will check about the htop and report directly to the OpenWRT maintainer for it. Thanks for these precious precisions.
 
I get it :)

It seems that the busybox traceroute issue was already reported, but I will double check.
I will check about the htop and report directly to the OpenWRT maintainer for it. Thanks for these precious precisions.
I do not mean that you should report this to maintainers from OpenWRT... It is just a part of Makefiles.

Voxel.
 
I do not mean that you should report this to maintainers from OpenWRT... It is just a part of Makefiles.

Voxel.
Yes, I see that.
I have contacted Etienne CHAMPETIER who advised to report it in the OpenWRT GitHub under htop package, and from there, he will have a look. He asked me if through Entware I had access to releases 3.0.2, 3.0.3 and 3.0.4, probably to check in which one the problem was introduced. I told him no, but that I can probably compile these versions. We will see.

Thank you again :cool:
 
Some update:
I was able to build htop 3.0.2 from the official tarball release and it works.
Then I did the same with htop 3.0.3, and it shows the issue.

The problem appeared between versions 3.0.2 and 3.0.3.
I reported this to the OpenWRT issues report for package htop.

To be continued...
 
Latest news:
Mr CHAMPETIER tried htop on OpenWRT
Hi bolemo
I just tried with a RaspberryPI 2, so cortex A7 but same instructions ARMv7-A and everything work fine with latest OpenWrt snapshot / htop 3.0.5-1

So this issue seems more specific to Entware.
Does anyone else here experience the same bug with latest Entware’s htop?
 
Ok. solution found for htop:

It needs to be configured with this option: ac_cv_func_openat=no before compilation.

And it works :)

EDIT: Also, I was informed that it is now already fixed in Entware, and they think @Voxel will likely do the same for his next release.
 
Last edited:
So I was wondering why my wordpress keep complaining that "zip" PHP module is not installed or not enabled.

Today when I run "/opt/bin/php-cgi" command and this is what I found:

Bash:
root@R7800:/tmp/mnt/sda1/entware/share/nginx/html$ /opt/bin/php-cgi
PHP Warning:  PHP Startup: Unable to load dynamic library 'zip.so'
(tried: /opt/lib/php/zip.so (libnettle.so.7: cannot open shared object file: No such file or directory),
/opt/lib/php/zip.so.so (/opt/lib/php/zip.so.so: cannot open shared object file: No such file or directory)) in Unknown on line 0

I search across the web and found one solution written in Russian:

Bash:
/opt/bin/opkg list-installed|grep -v "libc "|sed 's/ - .*$//'|grep lib|grep -v libpthread|grep -v libgcc|xargs -n 5 /opt/bin/opkg --force-reinstall install

After waited for a while for the packages reinstallation, the "zip" PHP module not install or not enabled issue went away.
 

Support SNBForums w/ Amazon

If you'd like to support SNBForums, just use this link and buy anything on Amazon. Thanks!

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top