I am running a web server based on CentOS 5.8 and I need to upgrade my version of bind to make it PCI compliant. I'm currently running bind 9.3.6 and I need to have bind 9.9.8 or higher. I've tried yum update bind but apparently I already have the latest version according to yum. I did some Googling and I found an RPM file bind-9.10.2-1.el5.i686.rpm which looks like it would work but i don't know if it should try installing it or not. I think I would need bind-devel and bind-libs which I can get from the same site. Am I better off compiling from source? I know CentOS 5 is old but I'm trying to avoid reinstalling the whole server.
Installing binary rpm's from later versions of CentOS is unlikely to work: there are many changes since CentOS5.
Rebuilding a src.rpm locally is one way to see what issues there are.
Meanwhile, upgrading to CentOS6 (at least: CentOS7 uses systemd which takes some study) is often not a whole lot more effort than retrofitting something like bind, and will have other efficiencies. YMMV, everyone's does.
Related
We’ve been trying to get SikuliX 2.0.5 to run on a RHEL 8 system, and not having much luck.
We went through the instructions on this webpage:
https://sikulix-2014.readthedocs.io/en/latest/newslinux.html#newslinux
We started on RHEL 7, but the OpenCV shared library required a newer version of GLIBC than is standard on RHEL 7 (version ‘GLIBC_2.27’ not found (required by ~/.Sikulix/SikuliLibs/libopencv_java430.so)), so we moved up to RHEL 8. We had to build OpenCV (v4.3.0) from source because we could not find a java companion package for RHEL 8, which required quite a few other dependencies, but in the end we got it built with most options enabled, and installed as root on the system. We also got Tesseract installed via a package, as well as xdotool and wmctrl.
We are setting LD_LIBRARY_PATH to ensure that the OpenCV libs are picked up, and when we run with the “-v -c” options to the IDE, there are no obvious problems reported. It seems to believe it is moving the mouse, though we can see that it is not, and when we try to capture a screenshot, the “canvas” from which to capture is either uninitialized/garbage frame buffer memory, or a totally black screen. On rare occasions we have seen the actual desktop, but most times we do not.
Originally the system had 2 monitors, but was subsequently reconfigured to a single display system. We were originally running remotely over NoMachine, but have also tried running locally and observed no difference in behavior.
Any pointers or suggestions would be most welcome. Given that no error messages are being reported, we are out of ideas for how to proceed in debugging the problem. It appears that more native support is provided for Debian-based systems, but we’re attempting to validate a product which only advertises support for RHEL systems, so we’d prefer to get it working in this environment if at all possible.
I have installed Motioneyeos on a Pi Zero and on a Pi2, and it works like a charm in both of them. The control of the 2 systems can be unified on the web server of any one of them. The web interface is clear and allows to customize hundreds aspects of the program. Anything perfect so far.
Trying to understand a little bit how does it work, I have logged to one of the 2 Pis through ssh and I have checked that the OS is kept to a bare minimum. uname -r returns 4.19.65.
On the web interface of the app it is possible to check if the last version of the program is the one running. I have checked, and yes, I have the last version.
My question is: does it make sense to upgrade the OS components used by the program (apt update, apt dist-upgrade) even if the last version of the program is the one running in my Pis? Apt is not even installed, so the first thing would be to bring it there and install it, but I am afraid that if I update the OS, the program may stop to work...
I answer myself thanks to the feedback received in another forum. Motioneyeos is an embedded software based on Buildroot and as such it is not possible to install any package on it. To be able to install a package on Motioneyeos, Motioneyeos has to be installed on an full OS as Raspbian following the instructions in Motioneyeos web ==>>
I am a newbie to Xen and want to download it on my machine which currently has CentOS7. I have been researching and experimenting for a couple of days but can't seem to find a straightforward answer on how to install a fully functioning Xen on CentOS7. I tried using the workaround at http://www.lairdscomputer.com/Blog/tabid/62/EntryId/74/Installing-Xen-4-on-Centos-7.aspx, but it seems that some of the packages it uses might be outdated.
Is it even possible to install Xen on CentOS7, even if it is missing some parts to it? Would it be better just to go back to CentOS6.x so that I can install Xen4CentOS?
Thanks in advance! Any advice is appreciated!
Afaik RedHat therefore CentOS is not supporting Xen whoever in fact it is possible.
I am using xen4centos on my Centos7 server, it was installed along with official article from wiki:
https://wiki.centos.org/HowTos/Xen/Xen4QuickStart
It it is not, please provide some logs, errors etc
I learn that RHEL and Centos has the exact same source code. So my question here is if one application built for RHEL 6.4, could we just install and make it running on Centos 6.4 without any update?
And another question is RHEL and Centos will always use the same version number for the same source code, right?
Yes, software built on RHEL 6.4 should work as-is on CentOS 6.4 (assuming you don't depend on one of the relatively few RHEL-specific details, mostly to do with branding).
CentOS, except where necessary, uses upstream RHEL sources unmodified. When they need to modify a package they indicate that in the release notes and in the release field of the RPM in question.
I want to update my dedicate Centos Machine, which installed WHM. I want to know that if I do update the OS, there any changes on WHM. Like any default Setting changed? or any other Problems.
So please suggest me that right way of update Centos without any problems, because many websites hosted on this Server.
Thanks
Pankaj Gupta
Pankaj, because there are many changes in a major CentOS release you cannot upgrade from 5.10 to 6.5. You need to put together a migration plan. First install CentOS 6.5 on another server, configure httpd and php/passenger/whichever and test that each hosted website is working properly. Then you can reinstall and reconfigure everything.
Give CentOS 7 a try.
Upgrading is not possible and even if you try around doing some search you will end up wasting lot of time by fixing the dependency related errors etc. So its always better and safe to do a fresh install of CentOS 6.5 and then migrate the sites.