New answers tagged

0

Old question, but i stumbled upon the same question yesterday. Maybe it will help someone in the future. Here is my answer (tested with glusterfs 5.5) Create in /etc/glusterfs a file named "glusterfsd.vol". Add the following config for each volume to the file to configure the source ip. volume <your-volume-name>-server type protocol/server ...


1

Short answer is CDN. Take a look at CloudFlare as an example. https://support.cloudflare.com/hc/en-us/articles/205177068-How-does-Cloudflare-work- https://www.sitepoint.com/what-is-a-cdn-and-how-does-it-work https://www.imperva.com/learn/performance/what-is-cdn-how-it-works In a nutshell: CDN both absorbs the load from multiple remote hosts and provides ...


1

Here, our Docker container is reachable through docker0 (network: 172.17.x.x) From your browser, you're trying to connect to some 10.x.x.x address. Using a Browser on the Docker host Assuming you just want to connect from your browser to a locally-hosted container, then the easier would just be to reach your container address (172.17.x.x). In some cases, ...


0

Get a Linux LiveCD. If you only want to capture for a while, just download a Linux live image, write it to a DVD or USB thumb drive and boot from it. There are images exactly for such cases which will already contain Wireshark, but any will do. You can install software on a LiveCD session (it will just be gone on the next reboot). Boot the live session, ...


0

I could create a 3rd network and connect the two containers to eachother, but that seems overkill "… Containers can join multiple networks which allows you to provide fine grained network policy for connectivity and isolation …" Hence you can try solving this by introducing gateway CT that would be connected to both networks. Otherwise you'd need to ...


0

Yes, but that can block your LAN or vpn peer-to-peer communications depending on your setup. I'd rather suggest specifying an outgoing interface which is one connected to your server provider. Like this: export INET_IFACE = ethX iptables -A FORWARD -o $INET_IFACE -d 10.0.0.0/8 -j REJECT iptables -A FORWARD -o $INET_IFACE -d 172.16.0.0/12 -j REJECT iptables -...


1

Verify your remote named pipe connection string, for a named instance the format is \\<computer_name>\pipe\MSSQL$<instance_name>\sql\query https://docs.microsoft.com/en-us/previous-versions/sql/sql-server-2008-r2/ms189307(v=sql.105)?redirectedfrom=MSDN


1

10GBASE-T isn't downward comaptible (nor is any other Ethernet standard by itself). Twisted-pair Ethernet uses Autonegotiation where each side announces its capabilities (speed & duplex), so the link partners can agree on the best common protocol between them before PHY initialization begins. The process is detailed in IEEE 802.3 Clause 28. The protocol ...


-1

Google for: ms sql connect via tcp/ip site:microsoft.com One of the first links leads to: Configure a Server to Listen on a Specific TCP Port


1

You are very certainly running Docker. This can be infered by a few informations: the bridge names look like Docker bridge bridges having an IP set use classical Docker bridge IPs: /16 in the 172.16.0.0/12 space you are a developer every bridge MAC address start with 02:42, which designate a private MAC (fine) but if they were truely random they would have ...


0

You can watch monitor live using Status / Traffic Graph. Or you could use the ntop or BandwidthD package to monitor over.


0

Your service is only listening on the loopback interface (127.0.0.1). To make it available on other interfaces bind it to 192.168.1.12 or 0.0.0.0 instead (to have it listen on all interfaces).


1

Same as you, I struggled with veth, and failed to make it work. Then I found this, which says: Before MACVLAN, if you wanted to connect to physical network from a VM or namespace, you would have needed to create TAP/VETH devices and attach one side to a bridge and attach a physical interface to the bridge on the host at the same time, as shown below. Now, ...


0

Parse the pcap and index it according to your requirements. There are plenty of large-scale free-form search systems (such as Xapian, Solr, ElasticSearch, etc) which can be fed with all the interesting information you want to search on.


0

With current Debian based systems syntax has changed. The key is to use format of address with slash subnet address 192.168.129.134/17, and not to use netmask keyword anymore. The preferred method for Debian based system is to use networking configuration files. The alias method prefers to use such order: eth0, eth:0:0, eth0:1, ... So first alias to ...


1

If you use cloud, then definitely not. If you use on-prem kubernetes then definitely yes. The only disadvantage of this set up will be setting up proper node labels to devide them by datacentre and pod affinity to schedule pods properly to a target datacentre. More complicated set up will be sharing pods between datacentres. In this case you will need a "...


0

If I understand you correctly the short answer is no, it is not possible to use nodes from different regions in single cluster. However you can run clusters in multiple zones. Kubernetes 1.2 adds support for running a single cluster in multiple failure zones (GCE calls them simply “zones”, AWS calls them “availability zones”, here we’ll refer to ...


0

For the record, you can actually copy and paste actual files to and from the terminal session and in between terminal sessions on this server version.


1

BASE-K is backplane Ethernet. Are you sure that NIC is intended to be cabled external to a chassis? Plug both types of modules into a switch. A single module handles one medium type as fast as possible. Given the power and physical space restrictions, a module that talked over multiple mediums would be exotic. If it exists at all.


0

Not sure if this is what you want to achieve but you can install/enable SNMP client on VMs and hosts, and install a cacti/other monitoring server. Configure the cacti to monitor each NIC of the VMs and the NIC of the host where the VMs are connected. Monitor the network utilization host and compare it to VMs network utilization.


0

You don't need the machines to "see" each other to map a drive. If you have a shared folder on either machine then map it from the other machine with the net use command: net use X: \\computername\sharename /persistent:yes


1

With a naive routing, the packets from vlan1 would go to the switch, uplink to the router, the router would masquerade with its public IP address and then realize that the destination is on vlan2 and send it back to the switch to send to vlan2. That depends on how you set up the router. It should be able to route between private subnets without SNAT. ...


2

compatible? - it might have a problem with PXE boot but is extremely likely to work with anything else. supported? - not likely, neither by HPE nor by Dell. If you want full Dell support you'll need to buy a Dell branded NIC.


1

To fully understand packet flow through Linux iptables, you will need this diagram or something similar for reference. In the diagram, ping lives in the "local process" box in the application layer. With your current routing table, the system wil "think" only traffic addressed to 10.8.0.0/24 will be routed to the VPN tunnel interface; everything else must ...


0

You can run a program as a service but in your case, I would consider running Wireshark as a scheduled task with the right arguments.


0

The answer can be found here : https://docs.docker.com/v17.09/engine/userguide/networking/get-started-macvlan/#macvlan-bridge-mode-example-usage I needed to create a new network interface that is actually bridging with the macvlan : ip link add mac0 link enp3s0 type macvlan mode bridge ip addr add 192.168.0.199/24 dev mac0 ifconfig mac0 up Commands need ...


1

To me, on Debian 9, it helped just to reinstall ping: apt-get install --reinstall iputils-ping


5

No, a basic Bourne/POSIX shell (/bin/sh) cannot be expected to include any built-in facilities for TCP connections. See comparision of command shells in Wikipedia. The bash shell would have TCP and UDP client capabilities with a special handling of certain filenames: for example, using /dev/tcp/<hostname>/<port> in input/output redirection on a ...


1

When you want to implement failover on WAN connections, you need to configure interfaces with the correct metric/priority, and then manually switch over, or implement a ping/http connect automatic switch over... but also suitable static routes. When interface data1 goes down, both ipv4 and ipv6 IP's must change over/renew: ip a (newer tool) ifconfig -a (...


5

While I am not sure if this will work on I350 NIC I think the answer is in the ip-link man page: vf NUM specify a Virtual Function device to be configured. The associated PF device must be specified using the dev parameter. [--cut--] state auto|enable|disable - set the virtual link state as seen by the specified VF. Setting to auto means a ...


0

The Debian wiki mentions what may be the same problem, for Stretch and Buster. If your system has the file /etc/network/interfaces.d/setup (mine does not) you could try removing it. Otherwise, I like your idea of adding the bridge on startup. You could try in /etc/network/interfaces: auto br0 iface br0 inet dhcp pre-up brctl addbr br0 && brctl ...


0

The ICMP redirects are cached for the time specified by gc_timeout (/proc/sys/net/ipv4/gc_timeout). Note, however, that that the timeout starts from the last activity, so in some situations the fault can go on forever. I am not sure if all Ubuntu distributions have this functionality.


4

The host key fingerprints and the "Are you sure you want to continue connecting?" prompt is generated by your SSH client, not the remote sshd. No amount of configuration on the sshd server will change that. The setting you need to change in either your local personal ~/.ssh/config or in local system-wide /etc/ssh/ssh_config is StrictHostKeyChecking. If you ...


1

This may not be correct for your environment, but a network connection between the host and the guest should not use the physical network at all. The guest has it's own MAC address and IP Address. When the host wants to talk to the guest the network stack on the host needs to lookup the MAC address for the guest's IP and should identify that as being on ...


1

It's Akamai DNS 'magic' determining where to send the client to get the request. As I understand it (after having a discussion with an Akamai network architect at an IP-Peering meeting) Akamai's DNS uses shortest path routing (typically BGP) from the enquiring DNS client to determine where the closest location is and provides that IP Address. The DNS ...


0

The only thing you do is set up your DNS records to Akamai's specifications. You only put in the address and CNAME records that they give you. Akamai will handle the task of determining which edge router is closest to any particular user.


0

After some more investigation I finally figured it out. I had forgotten to enable forwarding in the kernel. sudo sysctl -w net.ipv4.conf.all.forwarding=1


1

I've not done this myself but it looks to me that for ESXi (which unlike vCenter has no db to store the properties) you need to add the following to your spec file { ... "InjectOvfEnv": true, "PowerOn": true, ... I found an example of this in Step 6 of https://www.virtuallyghetto.com/2016/04/slick-way-of-deploying-ovfova-directly-to-esxi-vcenter-server-...


0

Yes, this is a typical PAC file usage, check the below example code: function FindProxyForURL(url, host) { // If the hostname matches, send to the proxy. if (dnsDomainIs(host, "exampldomain.com") || shExpMatch(host, "(*.abcdomain.com|abcdomain.com)")) return "PROXY 1.2.3.4:8080"; // DEFAULT RULE: All other traffic, send direct. ...


0

My suggestion would be that you split your 10.2.0.0/24 network into 2 distinct networks 10.2.0.0/25 and 10.2.0.128/25, first one internal to server A, second one internal to server B. Then, you can add a route from server A to server B’s network: ip route add 10.2.0.128/25 via 10.1.0.253 and from server B to server A’s network: ip route add 10.2.0.0/25 ...


0

My suggestion would be that you split your 10.2.0.0/24 network into 2 distinct networks 10.2.0.0/25 and 10.2.0.128/25, one internal for each server, and then use IP in IP encapsulation to route the packets between those two networks. You can find details how to do this in https://www.tldp.org/HOWTO/Adv-Routing-HOWTO/lartc.tunnel.ip-ip.html.


1

It doesn't. In almost any case I have seen this delay, it is the applications (app that access the share) fault. Windows just maps the share to a drive, but a lot of apps try to access it the good old UNC way, which triggers the CIFS client to try it first. Word (2010) for example does not (voluntary) store files on R:\, for example, but on \server\...


1

@John Mahowald's suggestions are sound, especially logging the IPs that connect to 172.17.178.100 to determine which hosts are still using the old IP. Depending on your situation, it may be possible to tunnel the IP you require from the old subnet to the new one. This could be helpful if you are under a deadline to complete the move of the server to the ...


1

I would begin by ensuring server does not have a firewall running; it is behaving as if it has. netstat shows output from behind the firewall, so it might not be what you want. Next would be trying to access port 6607 from server itself, i.e. telnet localhost 6607 telnet to.server.ip 6606 (they behave differently) to verify it is not the service.


0

Loose ideas: Maybe setting up routing on l3 switch is enough Try to use haproxy as load balancer between networks Use second network interface ( or even up address on the same interface) and start project of redesigning of applications Set up second oracle and replicate ( just joking. Nobody buy second oracle. All we are learning).


3

Like many enterprises, Hetzner runs switch port security on its network switches, meaning that only authorized MAC addresses can talk to the network. This means that the hardware address of the NIC in your server can speak, but the virtual MAC addresses of your virtual machines cannot. Thus it does not work to bridge your VMs directly to the outside world. (...


0

I think your version is better 192.168.0.101:137 <=> 2.2.2.2:137 192.168.0.101:138 <=> 2.2.2.2:138 192.168.0.101:139 <=> 2.2.2.2:139 192.168.0.101:445 <= > 2.2.2.2:445 and then IDC will be able to connect through 2.2.2.2


2

Migrate all applications to use DNS, and document your findings. As you do not know all of them, this will be a discovery process. Configure a new different IP address on the host and put that in DNS as the service address, db.example.net or so. Communicate to all possible application owners the rename project, and that things will not work if no action is ...


0

They told me i need to use static configuration and it worked (after restarting network / rebooting). (Note that the IPs under that link are likely specific only to a Hetzner servers) Adjusting the configuration files was very difficult via Hetzner console (which does not accept Ctrl key, | > $ " ' etc.). These were the commands i had to run (will almost ...


0

I can't ask Anton Danilov with a comment yet, "If I want to allocate 50% of bandwidth to be available for contention but I want to divide the remaining 50% up amongst users equally, ensuring a guaranteed minimum, how would one go about doing that?"


Top 50 recent answers are included