It's documented that Puppet will auto-require certain resources, e.g. a (puppet-managed) directory that is a parent of a File resource, or a group resource referenced in a user resource.
If an auto-required resource has been declared as a virtual resource, is it necessary to explicitly realize that resource, or will the auto-require cause it to happen implicitly?
I have a daemon listening on a virtual IP address attached to a cloned loopback interface (lo1) on FreeBSD 11.x.
Inbound traffic for that daemon could arrive on several different physical interfaces, however running tcpdump on loopback only shows traffic that also originates from the loopback, and won't show traffic that is routed there from the outside. (NB: even for that to work you have to monitor the original lo0 i/f, not the cloned one).
Is there any way that I capture traffic destined for that daemon on the loopback without having to monitor each physical port separately?
Post processing of pcaps with tcpmerge would be unwieldy, and there's also some special capture code running that is neither amenable to offline merging nor re-writing to handle multiple interfaces at the same time.
On a benchmark lab system running Fedora Core 27 I've got Intel X710 10GE cards and 12-core Xeon processors, configured with 12 NIC queues and RX Flow Hashing based on both IP addresses and port numbers.
This results in uneven balancing between cores, and inconsistent performance results. I've got irqbalance disabled, with a 1:1 mapping from NIC queue to CPU core configure via /proc.
My application is UDP based so I'm not really bothered about the hashing. I can't generate enough entropy across the inputs to the hash function to get an even distribution of outputs, so I'd quite like to just try plain round robin instead.
Is there a way to disable RSS for UDP and get round robin whilst still maintaining separate queues? All of the links I've found that talk about disabling RSS also seem to disable multiple queues.
I need to configure Munin to monitor multiple instances of a server daemon that run on a single host.
I've tried creating a different named symlink for each instance, but unless I also pass an env var that's appended to each service name I get "Service collision" errors.
With the suffix appended, Munin works, but then I just get one very large set of graphs, but where I'd very much prefer to have them grouped by instance instead. Later, I'd also like to aggregate those graphs.
How can I configure this so that I get node.instance.service type grouping instead of node.service_instance?
I'm using a third-party plugin (https://github.com/shuque/bind9stats) on FreeBSD with the Munin 2.0.33_1 pre-built package.
I need to configure an SMTP server to accept any email to any address that might happen to reach it on port 25 and deliver it to a local mailbox (i.e. a spam sinkhole)
The default supported MTA on my O/S is Postfix, of which I have no experience. I used to do a lot of Sendmail and Exim, but that was years ago.
Can this be done with Postfix?
What's current best practice to capture full kernel stack traces on a Linux system (RHEL 5.x, kernel 2.6.18) that occasionally panics in a device driver?
I'm used to the "old" SunOS way of doing things - crash dumps get written to swap, and on reboot the dump gets retrieved in the local file system. man 8 crash refers to diskdump, but that appears to be unsupported. and/or deprecated.
I've played with kdump, but it's unclear whether I can get a stack trace from that. Triggering a panic via Magic SysRq didn't create one. It also seems wasteful to reserve so much memory (128MB) just for a kexec crash recovery kernel.
Following on from this question, how can I tell tcpdump to use a specific protocol analyzer for a particular port?
I'm doing some DNS development using a test server running on port 5053, but I can't figure out how to tell tcpdump to pretend that this is really DNS traffic.
I can't use wireshark - the packets aren't passing past a machine that's got a GUI.
I'm about to take the plunge and install Win7 RC1 on my desktop machine. However I want to preserve access to my old HDD with XP on it. I don't really want to use full dual-boot since that would leave my new Win7 install on the d:\ drive, which it wouldn't otherwise be.
Is it possible to use Virtualbox to boot from that old real drive instead of from an image?
If so, what are the implications for drivers, etc, since the "new" running OS will be using VirtualBox's virtual hardware and not the original hardware on which XP thinks it was installed.