IBM Bladecenter on ESXi Network Setup

Similar Questions

• Problems IBM BladeCenter S ESXI 5.0

  Hello

  I need help because it's been more than me that I me down with my IBM Bladecenter S and VmWare esxi 5.0.

  Configuration of our Lady:

  BladeCenter S 8886

  5 blades HS22 7870 with SAS (CIOv) connectivity cards

  2 modules NUMBER

  Of we have the option configured "Lun to search the block devices" SAS connectivity card on "Lun 0 only" because we have the boot on san for hypervisors.

  Pour policies manage the paths to the disks since the ESXi are configured on 'Fixed '.

  Operating system: ESXi 5.0 (boot on san)

  Two clusters are made between the 5 blades with a shared storage pour access to VMS and allow the migration to hot.

  The problem:

  The hosts lose access to LUNS on which virtual machines are located.

  Reduce the number of this type Hauptstr we noticed when insulating each LUN on a single host, the problem was much less, the activate option but we happens the HA to pay and to make migration hot.

  The problem seems to happen when the number of i/O disk becomes high.

  Attempted procedures:

  Update of firmware Modules NUMBER, we went on a updated version officially ESXi 5.0 (1.2.3.006)

  Update the firmware of the blades with Tmob, updating also card SAS connectivity.

  The certifier we did support and compatibility of our material by material support of IBM, at the request of the IBM Software support.

  Sending many bundle vmsupport, with newspapers and configuration, software to analyze support.

  Notre Dame configuration support for IBM Labs we have indicated what they.

  The problem was also occurred en ESXi 4 Update 1 version, but only 2 times and with less of VMs, so less i/o.

  We are waiting for a correct solution on the part of IBM since September 2011, we still have to intervene too often day and night on our servers because of these crashes that even if they are less scouring with the insulation of the luns, still prevent us from sleep.

  Thanks for your help, she is welcome!

  Hi Jeremy,.

  We have the same problem IBM Bladecenter S with VMware ESXi 5.0.

  Case is IBM and VMware.

• ESXi Network Setup?

  Hi people

  I have a question for you guys, maybe it's "s too bad for you, but I am quite unsure of what follows. If I have an ESXi server with an edge quadport NIC and I have a Cisco switch with spanning tree active what happens when I:

  Set all 4 ports on the switch, create a vSwitch standard on the ESXi and attach all 4 cards active network with political LB "based on the original Port ID.

  Based on my understanding he will distribute all my virtual machines in all of these NICs attached and all virtual machines are able to communicate (send AND receive) through these 4 vNIC/NICs. There are toggled a virtual machine which is pinned to a VNIC to an another VNIC when I unplug the specific cable.

  BUT

  My colleagues told me NOT! You will only be able to send data on all cards, but receive vNIC alone due to the MAC source/destination of the virtual machines respectively address the NIC in the host.

  Is this true? Exactelly what this policy process "origin port ID". Because there is additional one called "Mac-based..".

  Often, we set up systems like this:

  A server has 2 x quad port NIC. There is a vswitch with 2 cards attached (initially 1, first on a 2nd map) and we set up "based on the source port ID. Further we do NOT set a battery or an Etherchanne on switches connected to the NICs (via cross).

  We want to only have simpe LB and failover in the event of a path failure. It will work correctly (sending and receiving)? My colleagues told me that will not work as long as the switches are not stacked? Why should it not? My understanding of the VM switch vNIC another that is connected to an another NIC which even once, is connected to another standalone switch.

  When the wenn must configure "based MAC" or "HASH IP"? Well, I know when configuring "IP pole" then you must implement an Etherchannel between the host and the switch and set the IP hash strategy or you do not have the additional bandwidth.

  Why would we need to assign adapters Eve and not active? Which is used to deny the problems with loops in an environment no spanning tree?

  How to configure the network (amount of NIC, vSwitches, portgroup assignment, vmotion, etc. and switches)?

  Thank you very much

  What happens when I:

  Set all 4 ports on the switch, create a vSwitch standard on the ESXi and attach all 4 cards active network with political LB "based on the original Port ID.

  Based on my understanding he will distribute all my virtual machines in all of these NICs attached and all virtual machines are able to communicate (send AND receive) through these 4 vNIC/NICs. There are toggled a virtual machine which is pinned to a VNIC to an another VNIC when I unplug the specific cable.

  BUT

  My colleagues told me NOT! You will only be able to send data on all cards, but receive vNIC alone due to the MAC source/destination of the virtual machines respectively address the NIC in the host.

  Is this true? Exactelly what this policy process "origin port ID". Because there is additional one called "Mac-based..".

  It is quite simple, in the default configuration of "based on the original ID Port" each vNIC a VM has only a single active physical uplink at some point in time. This physical NETWORK adapter is used by the host to send as well as receive traffic for this particular vNIC (because your physical switch will learn the VM vNIC MAC on the currently active port only will pass the traffic through this link).

  The distribution is static and only changes when you add/remove/connect/disconnect rising physics or the vNIC (power/power on, disconnect/connect the vNIC). The behavior is the same for the standard vSwitches but also distributed with this policy. You can see the current mapping in the sight of esxtop (r) network:

  

  The route in source option MAC hash function is very similar, but instead of using the internal ID virtual port, it establishes a static mapping based on the MAC address of an Ethernet frame transmitted by a vNIC source. This approach also maintains a static table of MAC on your physical switch to prevent the beating of MAC. A unique vNIC VM traffic will use several ports, if the virtual machine uses several source MAC addresses. Who should never usually happen unless you do some fancy stuff of networking within the virtual machine and multiple virtual interfaces and it means also you must allow forged passes and MAC changes in security of port vSwitch group options. This article summarizes very well:

  Hostile coding: VMware: MAC LB hash function

  A server has 2 x quad port NIC. There is a vswitch with 2 cards attached (initially 1, first on a 2nd map) and we set up "based on the source port ID. Further we do NOT set a battery or an Etherchanne on switches connected to the NICs (via cross).

  We want to only have simpe LB and failover in the event of a path failure. It will work correctly (sending and receiving)? My colleagues told me that will not work as long as the switches are not stacked? Why should it not? My understanding of the VM switch vNIC another that is connected to an another NIC which even once, is connected to another standalone switch.

  Your colleague is wrong, you don't need any special configuration to the ID load balancing mechanism base port. It will work very well with the recovery, due to the static mapping simple explained above. It will be just like plugging a system of one switch on the other (in the same broadcast domain).

  When the wenn must configure "based MAC" or "HASH IP"? Well, I know when configuring "IP pole" then you must implement an Etherchannel between the host and the switch and set the IP hash strategy or you do not have the additional bandwidth.

  MAC based transmission is also static, but based on the source MAC and has nothing to do with etherchannel/LACP. See my explanation and article above.

  Why would we need to assign adapters Eve and not active? Which is used to deny the problems with loops in an environment no spanning tree?

  vSwitches form loops unless you do really bad things in a VM with multiple network cards to configure some ornithology within this VM operating system. The active settings / standby are basically just primary/secondary hierarchy in case you want the traffic to a group of particular port through a specific binding, unless a failover occurs.

  How to configure the network (amount of NIC, vSwitches, portgroup assignment, vmotion, etc. and switches)?

  Depends on. On a lot of factors.

  To summarize:

  -separate your network with VLAN slipped into a totally physical configuration

  -vMotion put on a non-routed private VLAN with a dedicated physical connection (or active / standby time of team settings ensure vMotion is not shared with other traffic except in a case of failover)

  -use physical rising as much as you want for a bandwidth

  -for IP (NFS, iSCSI) storage or FT, use dedicated as well physical uplink

• esxi Network Setup 4 and esxi 5.5 the same

  Hello

  We have a customer with the following network configuration (see picture). They run ESXi and vCenter 4.0.

  We have added a new separate system with ESXi and vCenter 5.5 and to set up the network, the same.

  My question is this service console, I can't to be able to set this up, it's that are not part of ESXi 5.5?

  If you see the other image, I configured 2 one for management and one for vMotion VMkernel Port, they will be on the same network.

  Would that be correct?

  I don't mind not the network adapters are not connected again... and another thing must use active active alle 4 cards?

  Old configuration

  

  New configuration

  

  Since the release of vSphere 5, he didn't there no Service Console more... the management can be performed via VMkernel port with active management traffic.

  About vmnic design, my recommendation is:

  vSwitch0:

  For vMotion PortGroup - vmnic0 Active and standby vmnic3

  PortGroup management - vmnic3 Active and standby vmnic0

  vSwitch1:

  PortGroup for VMS - vmnic1 Active and active vmnic2

• ESXi Network Setup 4 nic

  Hello

  I'm currently setting up a new environment of ESXi 4.1. I have 4 NIC on each host. I have seen many different discussions on the separation of vmotion/management. However, I do not think what is ideal with the current network adapter configuration because should I choose to use 2 network cards for management and redundancy of management.

  That's why I think, I separate the 4 network cards on two different switches and on each one I create two trunks (or to cisco etherchannel) while both of these trunks includes 50% of the EPS to a single host. Should I choose to spend 2 network cards on management (one for each switch) I no longer have the ability to trunk/etherchannel to the switch and going so lose the two redundancy on this specific switch and the additional bandwidth, I get by both trunking interfaces.

  Any ideas on this subject?

  / Kris

  The only way I could think it would be with a managed switch making the port tcp based QoS on, I'm pretty sure that the management and the vMotiontraffic will be on different ports for what are your servers.

  It will depend on your brand, a model of switch and how it built of UZ ACL / Classes but it should be fairly easy to build given the level of the managed layer 3.

  The question is whether he can do this level of control of frame/packet without too general.

• Please help with ESXi Network Setup

  Hello, I would like to know how to put on a range of network in a single network adapter,

  because when you put two machines using the same virtual

  NIC, an ex-adresse IP: 200.100.100.100 and the other with 200.200.200.200, one of

  machines does not. Thank you.

  You can use the Exchange on your vSwitch, but in reality it is not best practice and I would suggest your more NICs for resilience and security

  If you have found this device or any other answer useful please consider useful or correct buttons using attribute points

  Tom Howarth

  VMware communities user moderator

  Blog: www.planetvm.net

• The Network Setup Wizard error message.

  When you try to open the Network Setup Wizard, I get the following error message.
  "To open this file, (ie rundll32.exe), Windows needs to know what kind of program created."
  Clues?

  Mac

  Hi JohnB,

  Welcome to the Microsoft community!

  When you try to open the Network Setup Wizard, you get back to open this,(ie rundll32.exe) file, Windows needs to know what kind of program has created.

  To work with the issue, I suggest you follow the steps:

  Method 1:

  Visit the site:

  Cannot find the Rundll32.exe file when you open Control Panel

  http://support.Microsoft.com/kb/812340

  Method 2:

  Analyze the computer to remove all possible malicious content

  Visit the site:

  Microsoft safety scanner

  http://www.Microsoft.com/security/scanner/en-us/default.aspx

  Note: The data files that are infected must be cleaned only by removing the file completely, which means that there is a risk of data loss.

  Feel free to contact us for any question related to Windows .

• I'm trying to connect 2 through a router (network) computers. The Network Setup Wizard tells me to insert my Windows XP CD, but I can't.

  Configure the connection without my Windows XP CD?

  
  

  I'm trying to connect 2 through a router (network) computers. The Network Setup Wizard tells me to insert my Windows XP CD, but I can't. Is there something that I can get online? What should I do?

  Hi Humphrey2,

  I suggest you arrange for an installation CD of Windows XP which may be necessary to copy the files to configure the network connection.

  You can follow these links and check if the problem persists:

  In Windows network connection issues

  How to troubleshoot a network home in Windows XP?

  
  Reference: Overview of the Network Setup Wizard

  Hope the helps of information.

• Connect IBM BladeCenter to the existing network

  Hello world

  I have a BladeCenter chassis with 2 blades installed with nortel esx 3.5 and 2 ethernet switches layer 2/3 and I'm testing how best to link this to my network environment.

  I'm not a networking person, I just want to know how of uplink my blade going to my regular switches. My question is: what's the point of having 4nics by blade for balance so then I uplink my blade switches with a single 1 GB eth cable? What is the right way to do this?

  Another thing, I noticed the configuration of VLAN is the following: consider that I used two external switches to test, I used a sw to test the lan and I attached a cable to each blade switch this switch and everything seems fine. then I used the other external switch to simulate my DMZ, I joined a new cable from each of the blade goes to this switch and a port goes into the FORWARDING State while the other is in a State of DEADLOCK. now, I would like to understand why this is happening.

  Thank you for your help, these questions may seem very basic to you people, but as I said that I'm not a networking person, so be patient

  A common way to do this would be to spend 2 x 2GbE trunks by blade, one of each truck at a different base switch. Maximizes availability and increases the throughput. There are many other things to consider so I suggest that read you on the VMware doco.

  Dave

• ESXi network problem

  I have a strange problem of networking with ESXi, my infrastructure is described below:

  I'm under vCenter 5.5 on a VCSA (5.5.0 update 2d, Build 2442330). The environment is a mixture of ESXi 5.1 (especially U2) and 5.5U1a. We have 1 primary site with 5 sites all remote connected via GRE VPN tunnels, The VCSA is in the main site, each remote site has at least 1 ESXi server managed by the VCSA. Each remote site has independent internet so for each site, the basic router (which ends the GRE VPN) routes inside networks (essentially 10.0.0.0/8) through the tunnel and has a default route to the local firewall. ESXi boxes that are not configured with the router as the default gateway, none of the custom static routes have been added to ESXi.

  Two of my remote sites (a site has two servers ESXi, the other has a site) ESXi servers keep losing contact with vCenter. Nothing else on these networks is never a problem to talk to the main site, including talking to the VCSA. Three servers are different versions of 5.1. Here's what I've done for troubleshooting:

  1. ESXi can ping on virtually any host through the tunnel to the main site with the exception of the VCSA.
  2. I isolated my network management in ESXi to a physical NETWORK card and made a SPAN on the switch port, a capture of packets revealed that when the rattling of the VCSA ESXi put the packet on the wire with a destination MAC of the firewall. For all the other hosts on the main site of subnet ESXi sends the packet with a destination MAC address of the core router. Packet capture on ESXi itself (using tcpdump-uw) shows the same thing.
  3. The ARP tables on all hosts show correct MAC addresses for the router base and the firewall.
  4. Make a "vmkping - I vmk0 - N < core router > < VCSA >" succeeds (destination MAC is correct in this case).
  5. Add a static route to ESXi to the address of the VCSA force it to the top of base does not work, ESXi continues to send the packet to the Mac static routes these firewall have been removed after the failed test of workaround.
  6. Restart the host will solve the problem for a short time.

  Some info:

  These three server 1 is an IBM x 3300 M4 using Intel I350 GB network cards, the other two are R620s from Dell. The R620s each have 1 4-port Broadcom BCM5720-1 4 Intel 82580 ports. IPv6 has been disabled on all three hosts. Right now vmk0 is linked to an Intel NETWORK card on three servers, but barely I started to watch it, so I don't know if we saw the problem with vmk0 with a Broadcom.

  The Dells are inherited, so we have not done installs it but I reinstalled one of them yesterday (it was 5.1 U2 I reinstalled 5.1 U3) at the request of support for VMware. Support noted that the firmware and drivers were outdated and did not meet the HCL for those who have been updated when I reinstalled.

  IBM has been purchased and installed by our care last summer, firmware and drivers correspond to this LIST with the exception of the version of the BIOS that is a future release (it corresponds to the HCL for 5.5).

  I have a pension case, SR 15633456803. Initially, they tried to blame the network. Although it is clear to me that this is not a network problem, I put a case of Cisco in any way. Cisco has quickly threw the evidence I had gathered and made the same determination. The only way that this could be a network problem is with ProxyARP. ProxyARP is disabled on my firewalls, ESXi has correct ARP tables and when you talk to anything on the main network, in addition to the VCSA it sends the packet to the MAC address so clearly not a question of ProxyARP. At this point VMware said essentially that they are running out of ideas. I hope my reinstalled host stay online for more than a week, then I'll reinstall the other two hosts and call it a day if all goes well, but I'm preparing for the possibility that may not happen.

  We never seen anything like this or have any ideas?

  I understand the problem, ESXi does not correctly release the generated ICMP redirects. You can force the issue by restarting the vmk0 interface (or any port of kernel VM is your management interface) VMware KB: Internet control management protocol redirects, because a permanent fix I will disable ICMP redirects on my routers.

• Networking question in VMware Workstation / nested ESXi network configuration problem

  Hi guys,.

  I am trying to set up a virtual lab, I setup a domain, a vcenter server and up to 1 ESXi host controller.

  I am really struggling with the network aspect in establishing a vmnic for storing shared, please see the attached screenshots and a few questions:

  1. Is a vmnic a real network map? or vmnic correspond to "network adapters" that you put in place since the setting of your virtual machine?

  I followed a tutorial and implement 3 NICs on the ESXi host, in Bridge mode, I tried this in my configuration of vmworkstation and I lose the connection to my server vcenter to the ESXi host.

  See screenshot attached to how I configured my ESXi host, screenshot is called 'screenshot2esxihost network', I run got each of the maps defined on NAT network.

  

  See also the screenshot of how I have my configuration of network settings in my VMWare Workstation, screenshot is called "screenshot1. the tutorial does not show how it has its editor of the virtual network configuration, but it puts its ESXi host NICs in BRIDGED mode, and I'm guessing that you can set this time within the virtual network adapter. If someone could tell me what changes should I make in there and on the network cards in my ESXi host, which would be really useful.

  

  My domain controller, server ESXi hosts and vcenter are currently on the 192.168.86 range, which currently stood at NAT

  See also a screenshot from the vsphere client - I don't see a vmnic "vmnic0", my goal is to be able to add a "vmnic2" for a separate storage network.

  

  any help would be received with gratitude, I VCP510 review at the end of the month and really hope to be able to get my lab setup as soon as possible.

  The vmnic is the rising 'physical' of an ESXi host. So with the 3 network cards that you configured for the host in VMware Workstation you will see under "Network adapters" and you can use them to vSwitches in the section "network".

  André

• Basic question about ESXi networks

  Hi all

  It may be a very simple question for you but I am new to ESXi and I need your help.

  I have a configuration of ESXi envionrment like that.

  192.168.1.200 - ESXi Server

  192.168.1.50 - a Linux machine

  192.168.1.60 - a Windows 8.1

  ESXi server connects to a router and the network is 192.168.1.0/24.

  The topology just like that.

  

  My problem is,

  1. I have a Mac machine connect to the same router. There is a web interface built into the Linux box. Can I use the Windows 8 to go, but I can't use the machine to physical Mac do. The firewall is off turn in ESXi already

  2. How can I make the two machines able to access internet?

  Thank you!!

  After taking a second look at the configuration of the network, I have seen that you have configured a VLAN ID on the port group. Assuming you are using a router not supported (i.e. without ports trunk/tag), this may cause the problem. Please remove the VLAN ID to see if that solves the problem.

  André

• ESXi network redundancy

  Hi all

  I would like to ask a few questions about my design of the network.

  I use the small DELL blade solution. (Dell VRTX).

  Currently, we have two servers with essential M620 HA cluster installed with ESXi 5.5 blades more license.

  I have decided to dismiss more my network configuration.

  Each blade has two ports 10 GB switch LOM connected internally (module e/s inside the chassis of the blade) and also connected 2 ports PCI network cards.

  As an attachment, you can find my layout of network design.

  EXPLANATION OF THE LAYOUT

  *********************************

  NIC1, NIC2 = LOM

  NIC3 NIC4 = PCI CARD

  NIC1 + NIC3 = kernel Ports (NIC 1 - standby adapter)

  NIC2 * NIC4 = Standart (NIC2 - Standby adapter) ports

  VSWITCH1 = used for V-MOTION, MANAGEMENT, NETWORK (same as our network LAN, VLAN7)

  VSWITCH2 = branches with 3 MPLS network VLAN

  CONNECTION of RED = traffic vlan 7 geocoded

  CONNECTION BLUE = trunk (VLAN1, 2, 3)

  CONNECTION GREEN = trunk (VLAN 1,2,3,7)

  BB is two stucked switch node.

  I would like to ask if my drawing is correct.

  Thanks in advance

  Hi Roman,

  First of all, I would be divided management and vMotion traffic VLANS separated. I've seen some process of vMotion goes wild, I like to keep this seperated ;-)

  Thus, you could use NIC1 to management, with NIC3 mode sleep for this and use NIC3 for vMotion with NIC1 pending.

  Secondly, I recommend you to use NIC2 and NIC4 in active/active mode, if you use the second card for the VM network traffic load balancing and not to waste it in fashion watches.

  See you soon

  Tim

• ESXi-Networking

  I have ESXi server with 3 maps and I'm looking for advice for best practices to connect:

  My network environment: external ISP connection Cisco router.

  Two switch (vLan interconnection).

  Please can someone help!

  As you got only 3 network cards then I would say to keep 1 card 2 management and vMotion network and other NETWORK cards network for traffic from virtual machines as virtual machine network is more critical because they are real workloads of your installation.

• Mudbrick ESXi networking Visio templates

  Hi all

  I'm looking to find templates in Visio that I can use to show how our ESXi server is configured for networking.  I need icons/shapes to show groups/vmnic and distributed switches port mapped to physical/vmnic NIC.

  Thank you very much

  The only template used there is the NETWORK card, the rest can be easily build up to standard visio shapes and graphics manipulation.

  The NETWORK card is contained in the set of template that David has provided the link too.