Insallation Loops on fresh Rocky 8 and Rocky 9

Bug reports
1

  • What version of apnscp are you running?
    newest, its literally a fresh install

  • Which version of RHEL or CentOS are you using?
    Tried Rocky 8 and 9 Fresh installs from minimal iso.

  • Can you reproduce this bug reliably? Provide steps. If not, provide steps to the best of your recollection.

curl https://raw.githubusercontent.com/apisnetworks/apiscp-bootstrapper/master/bootstrap.sh | bash -s - -s use_robust_dns=‘true’ -s always_permit_panel_login=‘true’ -s php_install_ioncube=‘true’ -s dns_default_provider=‘powerdns’ -s apnscp_admin_user=‘theusername’ -s apnscp_admin_email=‘emaiaccount@redacted.tld’ -s mariadb_version=‘11.8’ -s pgsql_version=‘17’ -s spamfilter=‘rspamd’ -s system_hostname=‘domain.redacted.tld’ ‘-redacted’

Also tried Rocky 10 but that one fails with something about nsswitch.conf in the FST. Since i did not want to debug that, i tried older supported rocky versions. Just to also have that not work out

2

may be related to this:
ll::run(): package zlib' already installed under siteinfo’“, " 0. Error_Reporter::add_error("package %s' already installed under %s’", ["zlib", "siteinfo"])”, " [/usr/local/apnscp/lib/log_wrapper.php:72]“, " 1. error("package %s' already installed under %s’", "zlib", "siteinfo")”, " [/usr/local/apnscp/lib/CLI/Yum/Synchronizer/Install.php:99]“, " 2. CLI\Yum\Synchronizer\Install->run()”, " [/usr/local/apnscp/bin/scripts/yum-post.php:21]", “ERROR: CLI\Yum\Synchronizer\Install::run(): package zlib' already installed under siteinfo’”], “stdout”: “”, “stdout_lines”: }
2026-07-05 13:27:31,419 p=518326 u=root n=ansible | localhost : ok=636 changed=10 unreachable=0 failed=1 skipped=357 rescued=0 ignored=0

3

Which provider? I just launched an Rocky 10 platform on Wednesday, no issues with nsswitch.conf.

 package zlib' already installed under siteinfo

Resolved on EL9 :+1:

4

All of them are direct installs direct from the officoal isos (minmal variant) from the official rocky website. Direct on a vm provisioned on proxmox (12 host cores, 90gb ram, 900gb storage).

how about getting the zlib resolved on other versions?

5

It only affects 8 and 9, both resolved. 8 goes EOL in ~3 years, whereas 9 has ~6 more years ahead of itself. I don’t see a good reason for 8, including software features - Ghost requires gcc10 minimum now that is only available on EL9+.

I’d need to see the error to proceed further as my install was from a minimal ISO as well.

6

Wich one would you like to see. right now i still have the vm from rocky 8 looping. But i can spin up 9 again as well… When i tried to run a setup with the command shown before, first on 10 i ran into the nsswitch.conf file error and as i want to get some site migrations done and did not want to waste time, i descided to go with 9 instead to just get things moving along - when that did not i work, i tried 8, because i have a few of those on production - but both 8 and 9 just have the zlib issue instead.

7

This is resolved by commit #f8f76f40 as noted earlier in the thread.