elyograg's questions

I have a Dell R170 server with Ubuntu installed on it. It is in UEFI boot mode. The system boots properly.

I have configured /etc/default/grub with options that should make it display the grub menu and wait 5 seconds before booting the default option. These settings work for other other systems in BIOS boot mode, and for libvirtd VMs that are in either BIOS or UEFI boot mode. Every time I change /etc/default/grub I run sudo update-grub.

I have been unable to figure out how to get it to show the grub menu and wait for the configured 5 seconds. What have I overlooked? I've tried a LOT of things to try and fix it, so far without success.

This is the contents of /etc/default/grub:

GRUB_DEFAULT=0
GRUB_TIMEOUT_STYLE=menu
GRUB_TIMEOUT=5
GRUB_RECORDFAIL_TIMEOUT=$GRUB_TIMEOUT
GRUB_DISTRIBUTOR=`lsb_release -i -s 2> /dev/null || echo Debian`
GRUB_CMDLINE_LINUX_DEFAULT="splash reboot=pci delayacct --backround=/boot/grub/splash.png"
GRUB_CMDLINE_LINUX="elevator=none"

All servers are Ubuntu 22.04 and I am using the OSS artifactory apt repo from jfrog.

I have a VM with artifactory installed on it. For various reasons I need to replace that VM with a new one.

I installed artifactory on the new VM. I was encountering a problem, can't remember what it was. So I purged artifactory and installed it again. Now it does not go through initial setup (changing the admin password, etc). The libs-release and libs-snapshot repos are not present.

How do I get it to go through the initial setup again, where I change the password and tell it to use Maven?

CentOS 7 MariaDB 10.2.44.

X [root@psa95 ~]$ rpm -qa | grep Maria
MariaDB-server-10.2.44-1.el7.centos.x86_64
MariaDB-common-10.2.44-1.el7.centos.x86_64
MariaDB-compat-10.2.44-1.el7.centos.x86_64
MariaDB-client-10.2.44-1.el7.centos.x86_64
✓ [root@psa95 ~]$ cat /etc/redhat-release 
CentOS Linux release 7.9.2009 (Core)

Credentials supplied in /root/.my.cnf with sections for client, mysqldump, mysql.

X [root@psa95 ~]$ mysql
Welcome to the MariaDB monitor.  Commands end with ; or \g.
Your MariaDB connection id is 4875
Server version: 10.2.44-MariaDB MariaDB Server

Copyright (c) 2000, 2018, Oracle, MariaDB Corporation Ab and others.

Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.

MariaDB [(none)]> Bye
✓ [root@psa95 ~]$ mysqldump mysql
mysqldump: Got error: 1045: "Access denied for user 'root'@'localhost' (using password: YES)" when trying to connect

The mysql command works. The mysqldump command doesn't.

I have tried using -p to prompt for the password, and supplying username and password on the commandline. Same thing -- works with mysql, does not work with mysqldump.

Another CentOS 7 system with exactly the same MariaDB version and the same setup in .my.cnf works perfectly. I have compared the mysqldump binary from both systems, they are identical.

I've scoured Google and tried every single idea I could find. None of it worked. What else can I try? What other information is needed to troubleshoot? I know the password is good -- it works with the mysql command.

I had tftpf-hpa working on my Ubuntu server to send/receive images and configs from my Cisco devices.

Today it is not working. The service won't start, even if I try it manually. It does not tell me why it won't start. If I run the start-stop-daemon with strace I get a lot of output, but I cannot see what the problem is.

I tried switching to atftpd, but that core dumps when it tries to write a file that I have sent from a cisco switch ... something I do quite often and has always worked before.

I also tried to use the tftpd package, but syslog indicates that it is trying to send the file I have asked the switch to send, rather than receive it. The switch indicates "Permission denied" as the problem. To test with tftpd, I changed ownership on /var/lib/tftboot and all its contents to nobody:nogroup.

I installed tftpd-hpa on another server running Ubuntu 20.04 and that is working. So I am not dead in the water, but that is a secondary server.

Can anyone help me get a working tftp server on my primary Ubuntu server? If there is a way to troubleshoot these problems, please let me know.

I have an apache2 config on Ubuntu. Everything was working well until I tried to move all my canonical redirect virtualhosts into a single file instead of having them scattered across lots of other config files.

I arranged my config filenames so that I could strictly control which site was loaded first, and therefore control which of them was Apache's default site. This is the first two files in /etc/apache2/sites-enabled, sorted lexically as ls does by default:

elyograg@smeagol:/etc/apache2$ sudo ls -al sites-enabled
total 12
drwxr-xr-x 2 root root 4096 Oct  6 16:13 .
drwxr-xr-x 9 root root 4096 Oct  6 16:12 ..
lrwxrwxrwx 1 root root   51 Jul 22  2017 000000-unknown.REDACTED.org.conf -> ../sites-available/000000-unknown.REDACTED.org.conf
lrwxrwxrwx 1 root root   47 Oct  6 16:13 0000-canonical-redirect.conf -> ../sites-available/0000-canonical-redirect.conf

Today I added the 0000-canonical-redirect.conf file and I want it to be the second site config file that Apache loads. It used to be that it would load the 000000 file as the default site, but since I added the new one, Apache is choosing that file instead of the 000000 one like I expected it to.

I will be downloading and looking at the source code, but I was hoping that maybe people have run into this before and know for sure that it is a bug.

elyograg@smeagol:~$ dpkg -l | grep apache2
ii  apache2                                    2.4.52-1ubuntu4.1                            amd64        Apache HTTP Server
ii  apache2-bin                                2.4.52-1ubuntu4.1                            amd64        Apache HTTP Server (modules and other binary files)
ii  apache2-data                               2.4.52-1ubuntu4.1                            all          Apache HTTP Server (common files)
ii  apache2-utils                              2.4.52-1ubuntu4.1                            amd64        Apache HTTP Server (utility programs for web servers)
ii  libapache2-mod-dnssd                       0.6-3.2                                      amd64        Zeroconf support for Apache 2 via avahi
ii  libapache2-mod-passenger                   6.0.10-3build2                               amd64        web application server - apache2 integration
ii  libapache2-mod-wsgi-py3                    4.9.0-1ubuntu0.1                             amd64        Python 3 WSGI adapter module for Apache

EDIT: I tried what I hoped would be a more explicit naming, but it still loads the redirect file first.

elyograg@smeagol:/etc/apache2$ ls -al sites-enabled/
total 12
drwxr-xr-x 2 root root 4096 Oct  6 16:41 .
drwxr-xr-x 9 root root 4096 Oct  6 16:12 ..
lrwxrwxrwx 1 root root   50 Oct  6 16:25 _0000-unknown.REDACTED.org.conf -> ../sites-available/_0000-unknown.REDACTED.org.conf
lrwxrwxrwx 1 root root   46 Oct  6 16:41 111-canonical-redirect.conf -> ../sites-available/111-canonical-redirect.conf

I'm having a problem where putty's key forwarding (via pageant) to my gitlab server isn't working. I'm doing a putty session to my server, with ssh auth forwarding enabled. I see the effects of that forwarding in the environment variables on the remote system.

Here's a transcript from an attempt that is not working. This is in a putty window.

login as: elyograg
Authenticating with public key "elyograg-home" from agent
elyograg@smeagol:~$ cd /etc/haproxy
elyograg@smeagol:/etc/haproxy$ git pull
[email protected]'s password:             (at this point I did Ctrl-C)
elyograg@smeagol:/etc/haproxy$ env | grep SSH
SSH_AUTH_SOCK=/tmp/ssh-PaSFdm2LDk/agent.3816409
SSH_CONNECTION=192.168.1.222 58109 192.168.1.200 22
SSH_CLIENT=192.168.1.222 58109 22
SSH_TTY=/dev/pts/0

The "git pull" command should have authenticated instantly with no password via the key forwarding. This HAS worked in the past.

What's really weird here is that Eclipse and Git for Windows, which both utilize plink (part of the putty install), are working just fine to the same gitlab server with the exact same keys. The plink program uses the same agent forwarding mechanisms as putty. So I know that it's not a matter of using the wrong key. I had putty version 0.76 installed, which is the latest release version at this time. I installed a pre-release 0.77 version with no change in behavior.

I'm on 64-bit Windows 10, fully up to date. CPU doesn't support Windows 11. The server where I am trying this is 64-bit Ubuntu Server 20.04, also fully up to date, running on a Dell R720xd.

I tried to find a support resource for putty, but it looks like they don't have one. They've listed a couple of usenet groups for general ssh discussion.

I have tried to find an answer to this and I have come up empty.

I have a certificate from LetsEncrypt that I use for all my SSL needs. It is valid for all the DNS names that I am trying to use. Today I tried to get it working for MySQL, but MySQL insists on using its own self-signed cert.

MySQL version is 8.0.26-0ubuntu0.20.04.2 from Ubuntu packages.

I have created /etc/mysql/mysql.conf.d/zz-ssl.conf with these contents:

[mysqld]
ssl-ca=/etc/ssl/certs/local/mainchain.pem
ssl-cert=/etc/mysql/cert.pem
ssl-key=/etc/mysql/cert.pem
[system_default_sect]
MinProtocol = TLSv1.2

But the log says this:

2021-08-08T16:16:57.982003Z 0 [Warning] [MY-013746] [Server] A deprecated TLS version TLSv1 is enabled for channel mysql_main
2021-08-08T16:16:57.982227Z 0 [Warning] [MY-013746] [Server] A deprecated TLS version TLSv1.1 is enabled for channel mysql_main
2021-08-08T16:16:57.983344Z 0 [Warning] [MY-010068] [Server] CA certificate ca.pem is self signed.

I tried deleting /var/lib/mysql/*.pem ... but when I restart mysql, all those files come back with updated timestamps. MySQL seems determined to use its own self-signed certificate and not the one I have configured. I figure I am missing something that should be obvious but isn't.

The file /etc/mysql/cert.pem referenced in my config is owned by mysql:mysql and has 0600 permissions. It is a copy of the certificate file I use for everything else, it contains the server cert, the private key, and the LetsEncrypt issuing cert.

I'm trying to set up a slave MySQL server on another continent. I don't have enough space on the new server for the entire database that the master contains. Therefore I am trying to bring up this server without the three largest databases. Once I get the server working, I intend to drop those three databases.

The DB heavily uses InnoDB, with the innodb-file-per-table option, so it's not a single ibdata1 tablespace file.

I tried copying over the data directory, minus the directories for the three extremely large databases. This didn't work, because it couldn't load the innodb tablespace for the tables in the databases that I didn't copy.

Then I copied the rest of the files, but used dd to copy only the first 100MB of each .ibd file in those three directories, instead of the many gigabytes that the files on the master contain. With the innodb recovery option, I can start MySQL, but I am unable to drop the tables or databases at that point.

How can I successfully bring up this slave without the three large databases, so I can get replication going on a subset of the remaining tables?