I have been using the Zabbly kernel for Ubuntu 22.04.4 LTS for several months now and it works great but Ubuntu still wants to install its own, generic kernel and I don't want it.
Question one: does the system need the generic kernel (modules, headers) for anything?
Question two: if the answer to question one is no, how do I blacklist the installation of new generic Ubuntu kernels?
I want to prevent the installation of any new Ubuntu kernels.
https://hub.docker.com/r/internetsystemsconsortium/bind9
When the author of this documentation says,
You need to properly mount the following volumes:
/etc/bind - for configuration, your named.conf lives here
/var/cache/bind - the working directory, e.g. options { directory "/var/cache/bind"; };
/var/lib/bind - this is usually the place where the secondary zones are placed
/var/log - for logfiles
do they mean we should create a directory tree in the directory where our Dockerfile lives?
Below this under the title "Authoritative DNS server" they have some config file syntax. I can't tell if it's intended for the named.conf.local or named.conf.options file. What is your guess?
Thanks.
Or put another way, when I see @OWNER in a nfs4 ACL does this @OWNER refer to a local Unix/Linux user?
I need to make an AD user (without a corresponding local account) the owner of a directory on a CentOS server.
When I say "without a corresponding local account" I mean the target directory is owned by a local service/app account that is not bound to any network (AD/LDAP/Kerberos) identity.
Thanks.
I have a tiny (three worker node, one controlplane) kubernetes cluster running and I want to configure portworx for storage. All my hosts are kvm virtual machines running Ubuntu 20.4.3 LTS with 5.4.0-73-generic kernel. I used https://central.portworx.com/specGen/wizard to generate a portworx 2.8.1.2-c9b840b (OCI) spec file and applied it to my cluster. I chose to not configure a dedicated kvdb drive on these hosts although each host does have a sdb intended as the portworx drive. (These sdb drives are not mounted.)
When I run pxctl status I see:
NODE NodeStartFailure 3a5bce08-b8ed-4fca-93a2-435a8d6e512b ALARM 1 Jan 9 18:31:45 UTC 2022 Jan 9 18:31:45 UTC 2022 > Could not find any available storage disks on this node. Node cannot act as a kvdb node as no storage device found. Waiting for other nodes to start kvdb. > If this is a storage node please add disks and restart Portworx.
NODE InternalKvdbSetupFailed 907d5a2c-297d-4587-87f8-31d539cff440 ALARM 1 Jan 9 18:31:10 UTC 2022 Jan 9 18:31:10 UTC 2022 failed to setup internal kvdb: failed to provision internal kvdb: timeout waiting for other kvdb nodes to start: Could not find any available storage disks on this node. Node cannot act as a kvdb node as no storage device found. Waiting for other nodes to start kvdb. If this is a storage node please add disks and restart Portworx.
I would rather use etcd instead of kvdb. Does anyone care to comment? Also, is the portworx operator intended for kubernetes?
oops. I just noticed this on the portworx website:
PX-Essentials support only built-in KVDB.
External ETCD is accessible in PX-Enterprise
Thanks.
I can't understand why the VMs I create using virt-install & kickstart do not have networking that can be accessed reliably from foreign hosts. The VMs I created using virt-manager can be logged into using ssh, are pingable etc. but the kickstarted ones are inaccessible.
Even if I virsh edit the kickstarted VM and change to the change did not survive reboot. kube2 was created using virt-install and kickstart. The other two were created with virt-manager.
[root@kraken msh]# virsh dumpxml kube2 |grep bridge
<interface type='bridge'>
<source bridge='br0'/>
[root@kraken msh]# virsh dumpxml kube1 |grep bridge
<interface type='bridge'>
<source network='host-bridge' bridge='br0'/>
[root@kraken msh]# virsh dumpxml kube0 |grep bridge
<interface type='bridge'>
<source network='host-bridge' bridge='br0'/>
What syntax do I use in my virt-install script to make sure the source network (host-bridge) is built? My virt-install --network syntax has varied so much I don't even want to post it here. Sometimes I'm able to ssh to the IP of the VM but not the hostname.
Do I need "--bridge br0" in my virsh-install syntax or something else?
I have to run a firmware update script that lives on the first hard drive of an Intel server board. I must use the UEFI shell to run this script, according to the Intel docs.
(Our servers have no external USB ports, so please don't suggest this.)
I have all the binary firmware blobs including the update script in /moo on blk0.
How do I mount the drive so I can run the script or do I need to mount the drive at all?
I'm trying to use the CentOS-8-x86_64-1905-dvd1.iso as the http source for my PXE/kickstart install.
I'm struggling to understand why the installer is looking of squashfs.img and a directory named LiveOS, neither of which exist in the above iso image.
Should I be using the CentOS-8-x86_64-1905-boot.iso instead? I followed https://docs.centos.org/en-US/8-docs/advanced-install/assembly_preparing-for-a-network-install but there was no mention made of which iso to use.
Apparently CentOS 7.3's installer had a screen where you had the option of telling Anaconda to use a particular "security profile" for the install.
Is there any way to invoke this during an install of CentOS 7.6? Perhaps there is a kickstart syntax.
I have variables set up in the %pre section that look like this:
%pre
VOLGROUP=$(lvdisplay |grep VG | awk '{print $3}' |head -1)
vgchange -a y $VOLGROUP
ROOTVOL=$(lvdisplay |grep root | grep Path | awk '{print $3}')
HOMEVOL=$(lvdisplay |grep home | grep Path | awk '{print $3}')
SWAPVOL=$(lvdisplay |grep swap | grep Path | awk '{print $3}')
%end
Are those variables accessible to Anaconda? So far they seem to be ignored. Here's how I would use them.
logvol /home --fstype="ext4" --noformat --useexisting --name="$LVHOME" --vgname="$VGNAME"
I'm running the CentOS 7.6 installer from a USB thumb drive using a kickstart file. I'm installing to a system that has only one disk.
In the bash shell of the installer, my system's hard drive shows up as /dev/sdb. The installer sees the USB drive it booted from as /dev/sda. Should I be telling Anaconda to use /dev/sdb? Seems like a bad idea.
I have a kickstart file that I generated while doing a manual, graphical install. I'm trying to use this kickstart file while booted to CentOS-7-x86_64-LiveGNOME-1810.iso and running
/usr/bin/liveinst -C --kickstart=/root/mykickstart.ks
but I'm having no success with getting Anaconda (liveinst) to recognize my instructions regarding partitioning of the disk.
Has any of you ever got this to work? Is kickstart the right path? What about these https://anaconda-installer.readthedocs.io/en/latest/boot-options.html# commands?
I'm tasked with 'upgrading' CentOS 6.x to CentOS 7.6 on a machine that has no connection to any CentOS repo, so this would involve overwriting the root logical volume and boot partition. I need to leave the home logical volume untouched. Ideally my user would click the "Install to disk" launcher on the desktop and not have to answer any installer questions but for now I'd just like to figure out how to communicate effectively with Anaconda.
Again when I say "upgrade" I really mean "fresh install" while leaving /home intact.
thnx
Target system is CentOS 6.x but resizing operations will be done in %pre section of CentOS 7.6 kickstart.
I have already run resize2fs -M /dev/mapper/centos-home successfully.
If I mount the volume, df says it's 100% full. This is good.
Now I need to reduce the size of the logical volume to match the size of the ext4 filesystem. I don't want to use rough estimates. I also don't need to conserve every possible extent. Within a >= 1GB range is fine. i.e. volume can be reduced to 1GB >= filesystem size.
The target volume does contain data I can't lose. I don't have shell (or any other) access to the system. The system has no CentOS repos available so I cannot upgrade it. I don't even have remote hands I can talk to. I'm using a different computer to build and test my install ISO on.
My %pre syntax/script needs to safely reduce /dev/mapper/centos-home so I can create a new root volume for CentOS 7.6 install. Yes, this entire proposition is nuts.
I have a project that requires me to shrink the root and home volumes on an existing CentOS 6 install to make room for a CentOS 7 install on a customer's bare metal host. Yes, I know what a bad idea this is. We ought to be shipping our customer a brand new, fully configured server.
Anyhow, I'm wondering if there's any magic utility that can scan unmounted disk volumes and report back how much free space or (or what percentage) can be used when I eventually run lvreduce. Anaconda must do this but how to I use anaconda in the %pre section of my kickstart config file?
Thanks.
From my ceph (mimic release) administrative node I ran
ceph-deploy osd create --data /dev/vdb ceph0
against a bare metal ceph node and it worked without error.
Now I run the same command against a virtual ceph node (with an unpartitioned qcow2 disk attached) and the command times out. I see no evidence of any progress having been made on this virtual machine. The target VM sees its vdb disk and virsh shows the disk as being properly attached.
The data disk on the bare metal machine is 500GB while the virtual disk on the virtual machine node is only 30GB. Could this massive size difference be the problem? Should I try using a raw .img disk instead of qcow2?
Thnx.
We have Dell DRM (not to be confused with digital rights management) installed in a VM that our production servers can reach. We have used DRM to download a catalog and the required update binaries ( firmware, BIOS etc.) for our hardware. Can we use DRM to push updates onto servers running CentOS Linux or do we need to use Dell dsu to pull them from our DRM local repo?
Thanks.
For some reason the OpenBSD 6.2 installer needs the base packages to be in "6.2/amd64" but these directories are not found by the installer. I think I might need to mount a disk partition manually but I can't figure out which one. Any clues would be appreciated. Burning CDs in 2018 seems silly.
When I attempt to pkg_add screen or any other pkg, I see:
ftp://ftp3.usa.openbsd.org/pub/OpenBSD/4.8/packages/i386/screen-4.0.3p1.tgz: unsigned package (signify(1) doesn't see old-style signatures)
ftp://ftp3.usa.openbsd.org/pub/OpenBSD/4.8/packages/i386/screen-4.0.3p1.tgz: unsigned package (signify(1) doesn't see old-style signatures)
ftp://ftp3.usa.openbsd.org/pub/OpenBSD/4.8/packages/i386/screen-4.0.3p1.tgz: unsigned package (signify(1) doesn't see old-style signatures)
ftp://ftp3.usa.openbsd.org/pub/OpenBSD/4.8/packages/i386/screen-4.0.3p1.tgz: unsigned package (signify(1) doesn't see old-style signatures)
ftp: local: -: Broken pipe
421 Service not available, remote server has closed connection.
Can't locate object method "errsay" via package "OpenBSD::PackageRepository::FTP" at /usr/libdata/perl5/OpenBSD/PackageRepository.pm line 845.
/etc/pkg.conf contains: installpath = http://ftp5.usa.openbsd.org/pub/OpenBSD/%c/packages/%a/
I have this mount on my nginx server:
/home/msh/CentOS-7-x86_64-DVD-1503-01.iso on /usr/share/nginx/www/iso type iso9660 (ro,relatime)
so the root of the iso file system is one step in from the nginx document root. So when I load http://192.168.10.2/iso in a browser, I see:
../
EFI/ 27-Mar-2015 22:34 -
LiveOS/ 27-Mar-2015 22:34 -
Packages/ 31-Mar-2015 23:42 -
images/ 27-Mar-2015 22:34 -
isolinux/ 27-Mar-2015 22:34 -
repodata/ 31-Mar-2015 23:43 -
CentOS_BuildTag 31-Mar-2015 23:45 16
EULA 27-Mar-2015 22:36 215
GPL 27-Mar-2015 22:36 18009
RPM-GPG-KEY-CentOS-7 27-Mar-2015 22:36 1690
RPM-GPG-KEY-CentOS-Testing-7 27-Mar-2015 22:36 1690
TRANS.TBL 31-Mar-2015 23:50 2883
My question is, with the CentOS-7-x86_64-DVD-1503-01.iso distribution, what url do I use in /var/lib/tftpboot/pxelinux/default ? Thanks.
I want to be able to authenticate users (via ssh) to specific groups of hosts based on the users' LDAP netgroup membership. The LDAP server is dsee7. The distro is OEL 6.5 with openldap as the client.
The only account we don't want to authenticate using LDAP is root.
sssd.conf looks like this:
[sssd]
config_file_version = 2
services = nss,pam,autofs
domains = default
[nss]
filter_groups = root
filter_users = root
reconnection_retries = 3
entry_cache_timeout = 300
entry_cache_nowait_percentage = 75
[pam]
[domain/default]
auth_provider = ldap
ldap_id_use_start_tls = True
chpass_provider = ldap
cache_credentials = True
ldap_search_base = dc=e-durp,dc=com
id_provider = ldap
ldap_uri = ldaps://lvl1.lvs01.edurp.com/ ldaps://lvl1.lvs01.edurp.com/
ldap_tls_cacertdir = /etc/openldap/cacerts
The /etc/openldap/ldap.conf looks like this:
URI ldaps://ldap.edurp.com/ ldaps://ldap1.edurp.com ldaps://ldap2.edurp.com
BASE dc=edurp,dc=com
TLS_CACERTDIR /etc/openldap/cacerts
TLS_CACERT /etc/openldap/cacerts/ldap_ca.cert
TLS_REQCERT never
nsswitch.conf looks like this:
passwd: files compat
shadow: files compat
passwd_compat: sss
shadow_compat: sss
group: files sss
hosts: files dns
bootparams: nisplus [NOTFOUND=return] files
ethers: files
netmasks: files
networks: files
protocols: files
rpc: files
services: files sss
netgroup: sss files
publickey: nisplus
automount: files ldap
aliases: files nisplus
And /etc/pam.d/system-auth reads:
auth required pam_env.so
auth sufficient pam_fprintd.so
auth sufficient pam_unix.so nullok try_first_pass
auth requisite pam_succeed_if.so uid >= 500 quiet
auth sufficient pam_sss.so use_first_pass
auth required pam_deny.so
account required pam_unix.so broken_shadow
account sufficient pam_localuser.so
account sufficient pam_succeed_if.so uid < 500 quiet
account [default=bad success=ok user_unknown=ignore] pam_sss.so
account required pam_permit.so
password requisite pam_cracklib.so try_first_pass retry=3 type=
password sufficient pam_unix.so md5 shadow nullok try_first_pass use_authtok
password sufficient pam_sss.so use_authtok
password required pam_deny.so
session optional pam_keyinit.so revoke
session required pam_limits.so
session [success=1 default=ignore] pam_succeed_if.so service in crond quiet use_uid
session required pam_unix.so
session optional pam_sss.so
So I tried adding a netgroup to /etc/security/access.conf like so:
#- : ALL : ALL
-:@eeederp: ALL
and restarted sssd just in case . . .
My user belongs to the netgroup eeederp and is still able to get in using his LDAP password.
Any clues would be appreciated.
I want to buy a Cisco router to use on my home network. I'm just hoping to get a unit that is not too old and is not a SOHO model. I want to run the latest Cisco OS just to learn how Cisco does things. I would use it to divide my network into two or three IP networks, firewalling, etc.
I'm just asking for a model name/number.
Thanks.