New answers tagged

0

In theory, it looks like the IPV6ADDR_SECONDARIES should support a range format in initscripts-ipv6: 3ffe:ffff::1f-2a will set a range from "3ffe:ffff::1f" to "3ffe:ffff::2a" (docs) In practice, Redhat's NetworkManager doesn't appear to support that format.


0

you could just delete the ipv4 configured addresses with the default ip tools. Something like: ip address delete ip.ad.dr.ess dev ethx or enable it: ip address add ip.ad.dr.ess dev ethx you should know what the address is and the interface to which it is assigned, of course, but that is left as an exercise for the OP.


0

it's a bit unclear what you are trying to do but let me see. If you want to test the dns resolution of AAAA records you do not need to disable ipv4, you can just query the dns server for the AAAA record: dig -t AAAA yourdomain.com if the server has the record it will return it successfully with ipv4 as well. If you want to test the dns client with ipv6 ...


0

Integrate with a network configuration scheme compatible with your operating system of choice. Ideally, one with a well-defined API for doing this disable IPv4 task. For example, NetworkManager has many ways to set ipv4.method to disabled. Some are even user friendly. Deploying config files is also an option. That tends to require root privilege. And ...


2

With the network namespace feature of Linux since about kernel 2.6.29 it is possible to run a program in its own network space that can be constructed with its own links, addresses and routing and can also be connected to the main network space of the host through routing. This can be done with userspace programs without docker or other container frameworks. ...


0

A lot depends on the 'nice' factor (which can be set and modified for each process) and the 'nice' factor of all the other processes currently running and how many CPUs are available and the speed of the 'front' bus, the 'back' bus, the data bus, etc


1

Linux programs rarely use system calls directly. Usually there is a libc function with the same name. There is a standard way to override libc (and any other dynamic library) functions: You write a small dynamic library that redefines the function you want (e.g. bind). Inside this library you can access the previous function definition with dlsym(RLTD_NEXT, ...


0

after I nuked and paved a 5th time I did this: nmcli con show nmcli con delete <all interfaces but the one i was connected to> nmcli con add ifname br0 type bridge con-name br0 nmcli con add type bridge-slave ifname eno1 master br0 nmcli con up br0 Then I cheated and added the static ips to the br0 with nmtui and fired up libvirt... viola... working


0

Solved with $ sudo ethtool -s enp3s0 autoneg on


0

I solved the problem, finally. Problem was that I was using the IP from 777.777.7.777/30 as the main IP of 777.777.7.777 with a gateway of 777.777.7.778, but the main IP should have been 777.777.7.779 because 777.777.7.777 was the network ip, not the usable ip. That's why it wouldn't connect. That IP above is obviously not a real IP, but in sequence, a /30 ...


1

This is to differentiate between the host's name and 'localhost.localdomain' since they could be confused on hosts without known fixed IP addresses. Localhost is 127.0.0.1 but there are 16.7 million IPs in the reserved loopback IPv4 space (all of 127.*.*.*, 127.0.0.1 through 127.255.255.255, as denoted by the CIDR 127.0.0.0/8); all IPs in that space resolve ...


2

what directory do we need to share? This depends completely on your software. Are /usr and /home adequate? This depends completely on your software. or do I need also others like /lib, /lib64, /opt ...? This depends completely on your software. Also, what options for NFS mount to use? This depends completely on your software.


2

Actually this is the ouput from the comman ip a list lo my debian 10 buster: 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host ...


4

Like most performance problems, it is complicated. How much bandwidth each task gets is a complex interaction between many things, at different layers of the network stack. Even without shaping. An incomplete list: CPU scheduler for when tasks (and driver interrupt handlers) can get on CPU How fast the tasks get their data, possibly limited by bottlenecks ...


2

Have you noticed that on a crowded, but not completely stuck, highway with multiple lanes some cars and lanes will actually be temporarily faster than others? Nobody is explicitly controlling the flow of cars and assigning "bandwidth" to these and similarly Linux is not explicitly allocating bandwidth between processes. What you see is mainly an ...


0

You define HTB's default to use classid 1:10, and then define only 1:1. 1:1 is never used, so there is no limitation done. Either define default to 1, to reach classid 1:1, or add a filter to selectively choose when to reach 1:1. So either rewrite it as: sudo tc qdisc add dev ifb0 root handle 1: htb default 1 sudo tc class add dev ifb0 parent 1: classid 1:...


0

To get A or B to see each other's LANs, opening the firewall to allowed traffic is the easy way. If your organization has a use case for connecting these networks, the firewall policy should be changed. Could be implemented in several ways: a VPN (IPSec, OpenVPN, WireGuard), some SD-WAN controller that manages VPN links for you, or perhaps direct from the ...


0

If you need to transfer a lot of data, i would suggest using Syncthing. By default it will use public relays if no direct connectivity between hosts. Protocol and transfer are encrypted, but if you're still worried, you could run your own relay on C, and setting it on A and B.


0

No you are not doing something wrong. NAT connections are tracked in the connection tracking table within the kernel. You can view this table at any moment in time using the conntrack command (conntrack -L is a good start) - you may need to yum install conntrack-tools if it's not available on your system. As I have previously explained here, when it comes ...


0

Do both to a test host and see what happens. As IP packets are higher level than either of them, your connectivity may be impacted. Unless your host has interface level redundancy. Administratively down is a thing for interfaces, including Linux. From https://www.kernel.org/doc/Documentation/networking/operstates.txt Linux distinguishes between ...


0

For anyone following this thread... I also posted my question here, and got the answer. The syntax for moving a port between VLAN 100 and the default VLAN is: root@liux:~# ovs-vsctl get port dpdk1 tag [] root@liux:~# ovs-vsctl set port dpdk1 tag=100 root@liux:~# ovs-vsctl get port dpdk1 tag 100 root@liux:~# ovs-vsctl set port dpdk1 tag=[] root@liux:~# ovs-...


0

Solved the problem: $ firewall-cmd --get-active-zones $ firewall-cmd --get-zone-of-interface=docker0 $ nmcli connection modify docker0 connection.zone public $ firewall-cmd --zone=public --add-masquerade --permanent $ firewall-cmd --reload


0

This network really needs to be sorted out and renumbered into different networks. There are no "nice" solutions to this problem as the entire design of IPV4 is broken. You can add individual IP addresses to directly connected routing tables with commands (assuming VLANA connects to eth0 and VLANB to eth1) like ip route add 192.168.0.10 dev eth0 ip ...


0

You want to connect from the ESXi console to another system. Do you have opened the firewall on the host for this connections? Look in "Configuration -> Security Profile" if your connection is opened. e.G. SSH Client is normally blocked. Additionally you must have a static route on your ESXi host for the guest OS network if the gateway VM is not your ...


1

There's no general explanation, it's just the matter of following what happens in the routing settings. 10.251.0.1, is both a local router address and part of 10.251.0.0/16. When receiving a packet to a local address, the router matches the local table using the very first ip rule with lowest preference: 0, before the rule for table 10, and matches. ...


0

Here is best solution ip route get 8.8.8.8 | awk -- '{printf $5}' or echo $(ip route get 8.8.8.8 | awk -- '{printf $5}') in bash script you can declare main_interface and use anywhere as $main_interface main_interface=$(ip route get 8.8.8.8 | awk -- '{printf $5}')


0

I have solved it. I had the mistaken idea that vlan is added when moving from regular netdev to a vlan netdev, the opposite is the truth. vlan is added when the packet move from the vlan netdev to the regular one, it is removed when the packet is moving from the regular netdev to the vlan one. The point of the test was to varify the pc (1) did in-fact had ...


1

Only way you can try that is probably by using the bridge interface but it won't be helpful as much because it still lists as one interface. You might want to consider using iptraf command.


0

TL;DR There is no need of iptables, marks, nor to relax Reverse Path Forwarding/Filter. Rather than the ip rules you used which won't match for outgoing packets, use the commands as described in the LARTC's documentation link you provided: ip rule add from $IP1 table T1 ip rule add from $IP2 table T2 then everything will work fine. Long version Even if ...


0

I think you were close. Your destination port (--dport) should be 80 then you can use the colon notation for the forward specification iptables -t nat -A PREROUTING -d 192.168.1.10 -p tcp --dport 80 -j DNAT --to-destination 192.168.1.11:8080


3

This is a bug in the current implementation of iputils-ping. https://github.com/iputils/iputils/issues/193 This should show the complete time the command has been running. Comparation with Version 20161105 from debian stretch which doesn't have this bug: ping -c 4 1.1.1.1 PING 1.1.1.1 (1.1.1.1) 56(84) bytes of data. 64 bytes from 1.1.1.1: icmp_seq=1 ttl=...


1

This time field means the total time it took to send and receive echo packets including delay time. So the total time for your entire ping operation. But what I do not understand is that the value its showing in your statistics! It should be much higher than 5ms.


0

I solved my problem. This works for me: table ip nat { chain prerouting { type nat hook prerouting priority 0; policy accept; tcp dport microsoft-ds counter packets 1 bytes 52 dnat to :1445 } chain postrouting { type nat hook postrouting priority 100; policy accept; } chain output { type nat hook ...


1

Assuming you want to use a different IPv6 address for a one-shot job or a short period, you can simply add (or change) an IP address using the ip command eg: ip addr add 2607:a:b:c:1::55 dev eth0 valid_lft 120 preferred_lft 120 In this example you add an IP address to interface eth0 with a lifetime of just 2 minutes. You can also tell Curl to use a ...


2

Doh! I launched my browser from a machine that only had access to one of the two networks. No surprise it didn't work.


0

You can register and point an unlimited number of different domain names to the same (pair of) name servers. A single name server can host a nearly unlimited number of domains , each with their own unique records.


Top 50 recent answers are included