Package Details: dnsdist 1.9.3-1

Git Clone URL: https://aur.archlinux.org/dnsdist.git (read-only, click to copy)
Package Base: dnsdist
Description: Highly DNS-, DoS- and abuse-aware loadbalancer
Upstream URL: https://dnsdist.org/
Keywords: dns
Licenses: GPL-2.0-only
Submitter: rgacogne
Maintainer: rgacogne
Last Packager: rgacogne
Votes: 12
Popularity: 0.001432
First Submitted: 2016-05-08 13:58 (UTC)
Last Updated: 2024-04-05 11:57 (UTC)

Latest Comments

1 2 3 Next › Last »

rgacogne commented on 2023-10-12 07:08 (UTC)

The 1.8.2 release has no changes from 1.8.1 and is only intended as a rebuild against an updated libh2o to fix a security issue, so I'm waiting to see if h2o-2.2 can be updated because otherwise it's pointless to bump this package.

rgacogne commented on 2022-11-07 08:53 (UTC)

@kescherAUR 1.7.3 was only released to release packages for new build targets, and has no changes whatsoever relevant to Arch. I will eventually update to prevent confusion but the switch to OpenSSL 3.0.x seems to introduce some link issues with h2o-2.2 so I'll wait a bit for the dust to settle.

kescherAUR commented on 2022-09-24 11:52 (UTC)

If your build is failing with lots of errors from ld, you likely need to add options+=(!lto) to the PKGBUILD for h2o-2.2, rebuild and reinstall that package before building this one.

mystique commented on 2022-03-28 17:13 (UTC)

@RomLecat How much ram did you need to build? I have a small vps.. 1GB ram.. 978 literally.. thanks in advance.

RomLecat commented on 2021-12-02 14:46 (UTC)

Nice catch, indeed it was OOM. Built successfully with more RAM allocated, thanks!

rgacogne commented on 2021-12-02 13:46 (UTC)

Yes, I re-built this morning in a clean chroot. Did you check for out-of-memory errors?

RomLecat commented on 2021-12-02 13:27 (UTC)

Do you build on fully updated Archlinux ? I just updated it yesterday right before compiling dnsdist on a fresh VM.

rgacogne commented on 2021-12-02 08:48 (UTC)

Unfortunately it still builds fine for me, so I can't reproduce the error. That looks like a compiler bug, though.

As for h2o my preference would be to disable support for ruby, but I'm afraid I don't maintain it.

RomLecat commented on 2021-12-02 00:47 (UTC) (edited on 2021-12-02 00:47 (UTC) by RomLecat)

Looks like it's broken:

{standard input}: Assembler messages:
{standard input}:640288: Warning: end of file not at end of a line; newline inserted
{standard input}: Error: .size expression for _ZZN10LuaContext6PusherIFK12ComboAddressSt10shared_ptrI11DNSResponseEEvE4pushIZNS_18registerMemberImplIS4_S2_Z27setupLuaBindingsDNSQuestionRS_EUlRKS4_E53_EEvRKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEET1_EUlRKS5_E2_EENSt9enable_ifIXsrN5boost22has_trivial_destructorIT_EE5valueENS_12PushedObjectEE4typeEP9lua_StateST_ENUlSZ_E_4_FUNESZ_ does not evaluate to a constant
g++: fatal error: Killed signal terminated program cc1plus
compilation terminated.
make[2]: *** [Makefile:1396: dnsdist-lua-bindings-dnsquestion.o] Error 1
make[2]: Leaving directory '/home/romain/.cache/yay/dnsdist/src/dnsdist-1.6.1'
make[1]: *** [Makefile:1523: all-recursive] Error 1
make[1]: Leaving directory '/home/romain/.cache/yay/dnsdist/src/dnsdist-1.6.1'
make: *** [Makefile:1150: all] Error 2
==> ERROR: A failure occurred in build().
    Aborting...
 -> error making: dnsdist

Also, h2o requires manual build with Ruby 2.7 (symlinked to /usr/bin/ruby), which is not ideal.

1 2 3 Next › Last »