Roger Creasy's questions

I have several directories in an ext4 filesystem that I can't delete. When I run rm -rf on the directories ps aux | grep rm shows a status of D for the process. I have tried allowing it to run for several days.

The same directories have the same issue with ls. I have tried to mv and to rsync the directories to a different block device - these commands have the same issue. I can mv directories within the same block device.

Possibly pertinent - the drive did fill up. I was able to delete enough other directories to get the total used space down to 85%.

What can I try?

I was adding a user to my server via MySQL Workbench. Something went awry; the addition failed. Afterwards my root account could no longer access the DB via Workbench. (access denied)
I was able to connect as root from the command line mysql -uroot -p but root has no privileges.
This system has been in place and working, until now, for over a year

The log shows:
Found invalid password for user: 'mysql.infoschema@localhost'; Ignoring user
Found invalid password for user: 'mysql.session@localhost'; Ignoring user
Found invalid password for user: 'mysql.sys@localhost'; Ignoring user

I have no user with GRANT privileges.
How can I correct the invalid Password issue for these system users?

MySQL 8
Centos 7

I am trying to use pdfcrop in a bash script on Centos 8. I cannot find a package with texlive-extra-utils (which provides pdfcrop).

How do I make pdfcrop available on Centos 8?

I have Nginx, running on Ubuntu 16.04 with several virtual hosts set up. After adding a LetsEncrypt cert, using certbot, I can no longer access my site with www.example.com Going to example.com works perfectly with either protocol. If I go to www.example.com with either protocol, the server returns the default server block. Here is my server block in the Nginx configuration files:

server {

listen [::]:443 ssl ipv6only=on; # managed by Certbot
listen 443 ssl; # managed by Certbot
ssl_certificate /etc/letsencrypt/live/example.com/fullchain.pem; # managed by Certbot
ssl_certificate_key /etc/letsencrypt/live/example.com/privkey.pem; # managed by Certbot
include /etc/letsencrypt/options-ssl-nginx.conf; # managed by Certbot
ssl_dhparam /etc/letsencrypt/ssl-dhparams.pem; # managed by Certbot

    root /var/www/example/public;

    # Add index.php to the list if you are using PHP
    index index.php index.html index.htm index.nginx-debian.html;

    server_name example.com www.example.com;

     location / {
    try_files $uri $uri/ /index.php?$query_string;
}

location ~ \.php$ {
    include snippets/fastcgi-php.conf;
    fastcgi_pass unix:/var/run/php/php7.2-fpm.sock;
}

location ~ /\.ht {
    deny all;
}
}
server {
listen 80;

server_name example.com www.example.com;
return 302 https://example.com$request_uri;
}

I want the server to respond to https://www.example.com OR http://www.example.com the same way it currently responds to https://example.com

I have a backup system that uses mounts that are put in place in fstab. I had a failure today because the mount was not there. I want to have my script check for the mount, then mount if it is not present.
i.e.

if mountpoint -q /mnt/mountpoint; then                                                                                                                          
   echo "`date` /mnt/mountpoint is a mountpoint"                                                                                                               
else                                                                                                                                               
   mount /mnt/mountpoint                                                                                                                      
fi

Given the above, will mount use the settings that are in fstab, or do I need to specify them in my script?

I am not a systems person, please forgive my lack of knowledge.

I have set up FreeIPA 4.5 on Centos 7. I was able to successfully manage user login to a test server (Fedora 27). So, FreeIPA is "working" (in that capacity anyway).

I am trying to use FreeIPA to manage access rights for a Samba server (again, Fedora 27), in a test environment for now. I am following instructions on the FreeIPA site.

So far I have added the service to the ipa server via ipa service-add cifs/sambatest.my.realm (with my working domain). However, I had to use --force because DNS failed (I know this is possibly related, I could not get past this issue either). When I try to ipa-getkeytab -s ipaserver.my.realm -p cifs/sambatest.my.realm -k /etc/samba/samba.keytab on the Samba server, I get Failed to add key to the keytab

Can anyone give me some guidance? Please let me know of any other info I can provide.

I am trying to redirect all requests within a directory on a server to https. I have tried the following in a separate config (within the config.d directory), and in an .htaccess file. Neither work, but cause no errors.

<IfModule mod_rewrite.c>
    RewriteEngine On
    RewriteCond %{HTTPS} !=on
    RewriteRule ^ https://%{HTTP_HOST}%{REQUEST_URI} [L,R=301]
</IfModule>

When I tried in a config file the above was within a block.

EDIT
Here is my code in the config file:

<Directory "/var/www/html/ssl_dir/">
    Options Indexes MultiViews FollowSymLinks
    AllowOverride All
    <RequireAny>
        Require ip xx.xx.xx.0/24
        Require ip ::1
    </RequireAny>
    <IfModule mod_rewrite.c>
        RewriteEngine On
        RewriteCond %{HTTPS} !=on
        RewriteRule (.*) https://%{HTTP_HOST}%{REQUEST_URI} [L,R=301]
    </IfModule>  
</Directory>

I am a software developer trying to help with a network issue (letting you know I don't know what I am doing). We are a very small shop.

We have a Samba network with a dozen or so Linux servers, and maybe 2 dozen users. Network discovery stopped working last week. Users cannot browse to network resources. Most users are Windows 7, I have an Ubuntu box and a Fedora box with the same issue.

I can connect by manually entering resource locations (by name). Windows users can as well, but have to manually connect by IP address.

There have been no intentional changes to the network. We did have a power interruption last week. We have rebooted the router, and the switches.

Can anyone give me some guidance? Thanks for any help you can offer.

EDIT:

We still have this issue. We have one Mac user - Network Discovery works for him.

I am trying to set up key-based ssh access to a Fedora 25 box. I have done this on many Ubuntu servers with no issues. I get Permission denied (gssapi-keyex,gssapi-with-mic,password).

Here is my set up process:

1. (server) create /home/[user]/.ssh
2. (local) copy key.pub to remote /home/[user]/.ssh/authorized_keys
3. (server) chmod 700 /home/[user]/.ssh
4. (server) chmod 600 /home/[user]/.ssh/authorized_keys
5. (server) chown -R [user]:[user] /home/[user]/.ssh/

I ran sshd in verbose mode. But, I cannot interpret the output. I think this is the pertinent bit:

debug1: SELinux support enabled [preauth] debug1: ssh_selinux_change_context: setting context from 'unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023' to 'unconfined_u:unconfined_r:sshd_net_t:s0-s0:c0.c1023' [preauth] debug3: ssh_selinux_change_context: setcon unconfined_u:unconfined_r:sshd_net_t:s0-s0:c0.c1023 from unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 failed with Invalid argument [preauth]

Is something in SELinux my problem? Other places I should look?

EDIT

Here is a link to gist with the results of ssh -vvv: https://gist.github.com/RogerCreasy/527d7520ce356074f4737ced51fb886d

And, a link to a gist with sshd_config: https://gist.github.com/RogerCreasy/1105333c869a08193754410679e81335

I am trying to install Postfix on an Ubuntu 16.04 server; this is a web server. I need Postfix for send-only for fail2ban and logwatch.

All Postfix tutorials show that apt install postfix ends with dialog windows. When I run sudo apt install mailutils all I get is a completed script, no install (that I can tell).

How do I configure my install? (there is no /etc/postfix/main.cnf)

Sorry for the lack of support info. I do not know enough to know what to give you. Ask, and I will post what I can.

I am interactively editing my hosts file with a bash script. I need to insert a line in my hosts file. The line needs to be inserted on the first blank line after a specific comment. The content I am inserting is in a variable.

Here is what I tried ($NEWLINE is set previously):

sed -i '# Insert New DNS / Virtualbox/$NEWLINE/g' /etc/hosts

I get no error, but the line is not written. Any advice?

I am trying to copy files from my primary data storage server into an OwnCloud instance. The data storage server has successfully mounted the user's directory on the OwnCLoud server, via webdav. I can copy files from the data server to the OwnCloud server; I can mkdir on OwnCLoud from the data server. However, I cannot recursively copy a directory from the data server. I get, for each directory I am trying to copy: cannot create directory '/path/to/dir\ with\ spaces\ in\ name': Invalid argument

Here is the command I am using: cp -R /dir/* /mnt/point/

Both servers are Linux. However, there are windows file and directory names (with spaces) in the directory I am trying to copy. I think that the issue is with the directory name having spaces in it.

I am trying to push data from one server to another using rsync. I need to set the owner and permissions on the remote server. Here is my rsync:

rsync --chown=process:sambausers --chmod=D770,F770 -rptlvog /home/cloudserv/test.txt [email protected]:/Drive_D/docs/temp

All works, except the owner is not set - owner is set to cloudserv, which is the user rsync is sshing as. Can anyone tell me what I am misisng?

I have LogWatch running on a Fedora 20 server. Logwatch gives the following every night:

--------------------- Connections (secure-log) Begin ------------------------

**Unmatched Entries**

CROND: pam_env(crond:setcred): Unable to open env file: /etc/environment: No such file or directory: 199 Time(s)

crond: pam_env(crond:setcred): Unable to open env file: /etc/environment: No such file or directory: 398 Time(s)

---------------------- Connections (secure-log) End -------------------------

What could be generating these errors?

I am writing a script to tar a directory, then upload the tarball to S3. The directory is named based on a date, month-year. Here is my attempt:

 TARNAME=`date --date="$(date +%Y-%m-15) -1 months" +' %B-%Y'`

 tar cvfz /tmp/$TARNAME.tar.gz  /Drive_E/drive_i_bu/monthly/$TARNAME
 /usr/local/bin/aws s3 mv /tmp/$TARNAME.tar.gz s3://test-roger/

The above works, if I am not doing the variable substitution. WIth the variable, I get: tar: July-2016.tar.gz: Cannot stattar (child): /tmp/: Cannot open: Is a directory

Additionally, tar appears to try and dive deeper into the directory. Any ideas? Again, the above works without the variable substitution. So, the problem is not with the AWS CLI tool.

I have a Linux server running on a VMware virtual machine, with 4 virtual hard drives. After the box ran for a month, I added 2 of the 4 hard drives in the vSphere client; I need more space. I did this step a few weeks ago, then was pulled into another project before creating the file systems and setting up mounts. Now, I do not know which drive is which within Linux. I have /dev/sda , /dev/sda1, /dev/sda2, and /dev/sdb

How do determine which drives have existing data and which are the new? Or, how do I remove drives and start over (I know how to remove the drives in teh vSphere client, but not how to remove the references to them in Linux).

Here are the results of dmesg| grep sd:

[    1.361162] sd 2:0:0:0: [sda] 16777216 512-byte logical blocks: (8.58 GB/8.00 GiB)
[    1.361205] sd 2:0:0:0: [sda] Write Protect is off
[    1.361210] sd 2:0:0:0: [sda] Mode Sense: 61 00 00 00
[    1.361253] sd 2:0:0:0: [sda] Cache data unavailable
[    1.361257] sd 2:0:0:0: [sda] Assuming drive cache: write through
[    1.363223] sd 2:0:0:0: Attached scsi generic sg1 type 0
[    1.363398]  sda: sda1 sda2
[    1.363788] sd 2:0:0:0: [sda] Attached SCSI disk
[    1.364425] sd 2:0:1:0: [sdb] 1572864000 512-byte logical blocks: (805 GB/750 GiB)
[    1.364466] sd 2:0:1:0: [sdb] Write Protect is off
[    1.364471] sd 2:0:1:0: [sdb] Mode Sense: 61 00 00 00
[    1.364512] sd 2:0:1:0: [sdb] Cache data unavailable
[    1.364515] sd 2:0:1:0: [sdb] Assuming drive cache: write through
[    1.370673] sd 2:0:1:0: Attached scsi generic sg2 type 0
[    1.405886]  sdb: unknown partition table
[    1.406228] sd 2:0:1:0: [sdb] Attached SCSI disk
[    4.493214] Installing knfsd (copyright (C) 1996 [email protected]).
[    4.493849] SELinux: initialized (dev nfsd, type nfsd), uses genfs_contexts
[    5.933636] EXT4-fs (sdb): mounted filesystem with ordered data mode. Opts: (null)
[    5.933649] SELinux: initialized (dev sdb, type ext4), uses xattr
[    6.099670] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: (null)
[    6.108488] SELinux: initialized (dev sda1, type ext4), uses xattr

Output from fdisk -l

Disk /dev/sda: 8 GiB, 8589934592 bytes, 16777216 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: dos
Disk identifier: 0x000dfc09

Device    Boot     Start       End  Blocks  Id System
/dev/sda1 *         2048   1026047  512000  83 Linux
/dev/sda2        1026048  16777215 7875584  8e Linux LVM


Disk /dev/sdb: 750 GiB, 805306368000 bytes, 1572864000 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes

Disk /dev/mapper/fedora_dataserv-swap: 820 MiB, 859832320 bytes, 1679360 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes

Disk /dev/mapper/fedora_dataserv-root: 6.7 GiB, 7201619968 bytes, 14065664 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes

I am trying to do a local rsync, from a mount point to a local folder. I need to set the owner, group, and permissions to specific settings. Here is what I am using:

 rsync -rtlv --chown=process:sambausers --chmod=D770,F770 /mnt/owncloud_mnt/Engineering/ /Drive_D/docs/Engineering_test

I end up with permissions 760 on both directories and files, and root:root on ownership (rsync is run as root).

What am I doing wrong?

TIA