Cisco UCS network uplink on aggregation layer

Hello Cisco Community,
we are using our UCS (Version 2.21d) for ESX Hosts. Each host has 3 vnics as follows:
vnic 0 = VLAN 10 --> Fabric A, Failover Fabric B
vnic 1 = VLAN 20 --> Fabric B, Failover Fabric A
vnic 2 = VLAN 100 --> Fabric A, Failover Fabric B
Actually UCS is connected to the Access Layer (Catalyst 6509) and we are migrating to Nexus (vPC). As you know, Cisco UCS Fabric Interconnects can handle layer2 traffic itself. So we are planning to connect our UCS Fabric Interconnects directly to our new l3 nexus switch.
Does anyone have connect the UCS directly to l3? Do we have to pay attention to something? Are there some recommendations?
thanks in advance
best regards
/Danny

we are using ESXi 5.5 with dvswitche (distributed vswitch). In our cisco ucs powerworkshop, we discuss pros and contras of hard- and softwarefailover and we commit to use hardwarefailover. It is very fast and we have no problems actually.
This is a neverending and misunderstood story: your design should provide load balancing AND failover. Hardware failover only gives you the latter.In your design, you just use one fabric per Vlan, what a waste !
And think about the situation of a failover on a ESXi host with 200 VM's; one has to send out at least 200 GARP messages, and this is load on the FI CPU. Most likely dozens or more ESXi server are impacted......
Cisco best practise: if you use softswitch, let it do the loadbalancing and failover, don't use hardware failover.
see attachment (the paper is not up to date,
For ESX Server running vSwitch/DVS/Nexus 1000v and using Cisco UCS Manager Version 1.3 and below, it is recommended that fabric failover not be enabled, as that will require a chatty server for predictable failover. Instead, create regular vNICs and let the soft switch send
gARPs for VMs. vNICs should be assigned in pairs (Fabric A and B) so that both fabrics are utilized.
Cisco UCS version 1.4 has introduced the Fabric Sync feature, which enhances the fabric failover functionality for hypervisors as gARPs for VMs are sent out by the standby FI on failover. It does not necessarily reduce the number of vNICs as load sharing among the fabric is highly recommended. Also recommended is to keep the vNICs with fabric failover disabled, avoiding the use of the Fabric Sync feature in 1.4 for ESX based soft switches for quicker failover.

Similar Messages

Cisco UCS 6120XP uplink considerations

HI,
I am working on a UCS design for a customer, I need to understand the physical connectivity of uplinks from 6120XP. Customer has 6509 in their environment with 48 x 1G eth modules. Unfortunately I cannot uplink 6120 using ethnet port to the LAN environment, Any suggestion or solution for followings:
- Can I connect/uplink 10G port from 6120XP to 1G port port on LAN switch probably 6509 or 2960.?
- Incase if Optical port is not available on 6509, can I use fabrci converter to convert the uplink from optical to eth, so that I could connect 6120XP with 6509?
Regards,
Kashif

Hi Kashif,
Yes you can connect the fabric Interconnect (FI) to 1G port of an uplink switch. The first 8 fixed ports can be configured to operate in 1 Gigabit Ethernet mode with the transceiver options specified for use with SFP-compatible ports using 1000BASE-T SFP GLC-T, you must specify the 1G rate of the port from the UCS manager.
I hope that I was in help to you!
Greetings,
Stanley

Windows 2008 R2 on Cisco UCS B200M networking problems

This is driving me completely nuts. Let me start by saying I am new to blade servers and Cisco UCS. I did take an introduction class, and it seemed straight-forward enough. I have a chassis with two B200M blades, on which I am trying to configure two Windows 2008 R2 servers, which I will eventually make Hyper-V servers. This is all in a test environment, so I can do anything I want to on them.
Right now I have installed W2008 directly on hard disks on the B200M hardware.
The problem is this: even though I think I've configured the network hardware correctly, using the Cisco VIC driver software, I cannot get networking to work in any reliable way. I cannot even get ping to work consistantly. I can ping my local server address, but I cannot ping my gateway (HSRP address). When I try, I get a "Reply from 10.100.1.x: Destination host unreachable (x being each particular server's last octet). I CAN, however, ping the individual IP addresses of the core switches. I can also ping some, but not all, the other devices that share the servers' subnet. There are no errors being generated, the arp tables (for those devices I can ping) look good, netstat looks OK. But I cannot get outside the local subnet...
Except when I can.
There are times when I can get all the way out to the Internet, and I can download patches from Microsoft. When it works, it works as expected. But if I reboot the server, oftentimes networking stops working. Yet another reboot can get things going again. This happens even though I've made no changes to either the UCS configs or the OS.
I cannot figure out any reason when it works at some times and not at others. I've made sure I have a native VLAN set, I've tried pinning to specific ports on the Fabric Interconnects. There is just no rhyme or reason to it.
Anyone know of where I can look? I'm very familiar with Windows on stand-alone boxes (although it's no longer my area of expertise), and I manage a global WAN (BGP, OSPF, Nexus 7k, etc.) so I'm no dummy when it comes to networking, but I am utterly stumped on this one.

The problem was this: while the NICs on the blade server are called vNIC0 and vNIC1, Windows was calling vNIC1 "Local Area Connection" and vNIC0 "Local Area Connection 2". So what I configured on UCS did not match what I was configuring in Windows. Completely, utterly ridiculous.
Anyway, networking is working now without any issues. Thanks for you suggestion; it did get me looking in the right direction.

Just like HP servers, does Cisco UCS C-Series Standalone supports Network Fault Tolerance ?

Just like HP servers, does Cisco UCS C-Series Standalone supports Network Fault Tolerance ? if yes then, where can i find this option
Windows :- 2k8 R2

Just like HP servers, does Cisco UCS C-Series Standalone supports Network Fault Tolerance ? if yes then, where can i find this option
Windows :- 2k8 R2

Nexus 1000v UCS Manager and Cisco UCS M81KR

Hello everyone
I am confused about how works the integration between N1K and UCS Manager:
First question:
If two VMs on different ESXi and different VEM but in the same VLAN,would like to talk each other, the data flow between them is managed from the upstream switch( in this case UCS Fabric Inteconnect), isn'it?
I created a Ethernet uplink port-profile on N1K in switch port mode access(100), I created a vEthernet port-profile for the VM in switchport mode access(100) as well. In the Fabric Interconnect I created a vNIC profile for the physical NICs of ESXi(where there are the VMs). Also I created the vlan 100(the same in N1K)
Second question: With the configuration above, if I include in the vNIC profile the vlan 100 (not as native vlan) only, the two VMs can not ping each other. Instead if I include in the vNIC profile only the defaul vlan(I think it is the vlan 1) as native vlan evereything works fine. WHY????
Third question: How it works the tagging vlan on Fabric interconnectr and also in N1K.
I tried to read differnt documents, but I did not understand.
Thanks

This document may help...
Best Practices in Deploying Cisco Nexus 1000V Series Switches on Cisco UCS B and C Series Cisco UCS Manager Servers
http://www.cisco.com/en/US/prod/collateral/switches/ps9441/ps9902/white_paper_c11-558242.html
If two VMs on different ESXi and different VEM but in the same VLAN,would like to talk each other, the data flow between them is managed from the upstream switch( in this case UCS Fabric Inteconnect), isn'it?
-Yes. Each ESX host with the VEM will have one or more dedicated NICs for the VEMs to communicate with the upstream network. These would be your 'type ethernet' port-profiles. The ustream network would need to bridge the vlan between the two physicall nics.
Second question: With the configuration above, if I include in the vNIC profile the vlan 100 (not as native vlan) only, the two VMs can not ping each other. Instead if I include in the vNIC profile only the defaul vlan(I think it is the vlan 1) as native vlan evereything works fine. WHY????
- The N1K port profiles are switchport access making them untagged. This would be the native vlan in ucs. If there is no native vlan in the UCS configuration, we do not have the upstream networking bridging the vlan.
Third question: How it works the tagging vlan on Fabric interconnectr and also in N1K.
- All ports on the UCS are effectively trunks and you can define what vlans are allowed on the trunk as well as what vlan is passed natively or untagged. In N1K, you will want to leave your vEthernet port profiles as 'switchport mode access'. For your Ethernet profiles, you will want them to be 'switchport mode trunk'. Use an used used vlan as the native vlan. All production vlans will be passed from N1K to UCS as tagged vlans.
Thank You,
Dan Laden
PDI Helpdesk
http://www.cisco.com/go/pdihelpdesk

Cisco UCS FC Direct Attach Question

We are looking at the Cisco UCS as a replacement for our existing servers and SAN switches. As I understand the fabric interconnect can replace our existing SAN switches and that we will still be able to zone the ports just like we do on our SAN switches today.
Can someone confirm how using the fabric interconnects as a replacement for our SAN switches will work? I read that the fabric interconnects have to be in switch mode for this to work. How does this affect the other connections we will have to our Ethernet network?
Thanks.

Q1: The 6120s run in NPV (N_Port Virtualization) mode which means it acts as a proxy for the blade vHBAs and doesn't participate in flogi (fabric logins). When a fabric switch runs in this mode it requires an uplink to a faber switch for flogi and zoning. Due to this you cannot direct attach the the 6120 to an array's front-end fiber ports.
Q2: Yes you can uplink the 6120 to a Brocade fabric as long as the Brocade supports NPIV.

ISCSI- Adding iSCSI SRs to XS booting via iSCSI [Cisco UCS]

Hello all-
I am currently running XS 6.0.2 w/ HF XS602E001 & XS602E002 applied running on Cisco UCS B200-M3 blades. XS seems to be functioning ok, except that i cannot see the subnet assigned to my SAN, and as a result, i cannot add any SR for VM storage.
The subnet and vNIC that i am configuring is also used for the iSCSI boot.
The vNIC is on LAN 2 and is set to Native
The SAN is a LeftHand P4300 connected directly to appliance ports (LAN2) on the Fabrics
Used the following commands during installation to get the installer to see the LUN.
echo "InitiatorName=my.lun.x.x.x" > /etc/iscsi/initiatorname.iscsi
/opt/xensource/installer/init --use_ibft
If i missed anything or more info is needed, please let me know.

Thanks Padramas,
I have 2 NICs that show up in XenCenter. Nic A is the vNIC associated with VLAN1 and provide the uplink to the rest of the network. The second NIC in XenCenter matches the MAC of the vNIC used to boot XenServer, so the NIC used for booting the OS is visible in XenCenter.
I have tried many different thing on my own to resolve this issue. I believe i have already tried adding a 3rd vNIC to the service profile for VM iSCSI traffic, but with make another attempt.
When configuring the vNIC for VM iSCSI traffic, do I only need to add the 3rd vNIC, or do i need to create both a vNIC and second iSCSI vNIC with Overlay vNic pointing to newly created (3rd) vNIC? My understanding is that the iSCSI vNICs are only used for booting, but I am not 100% sure.
Thanks again for the help!

Ask the Expert: Cisco UCS B-Series Latest Version New Features

Welcome to this Cisco Support Community Ask the Expert conversation. This is an opportunity to learn and ask questions about the Cisco UCS Manager 2.2(1) release, which delivers several important features and major enhancements in the fabric, compute, and operational areas. Some of these features include fabric scaling, VLANs, VIFs, IGMP groups, network endpoints, unidirectional link detection (UDLD) support, support for virtual machine queue (VMQ), direct connect C-Series to FI without FEX, direct KVM access, and several other features.
Teclus Dsouza is a customer support engineer from the Server Virtualization team at the Cisco Technical Assistance Center in Bangalore, India. He has over 15 years of total IT experience. He has worked across different technologies and a wide range of data center products. He is an expert in Cisco Nexus 1000V and Cisco UCS products. He has more than 6 years of experience on VMware virtualization products.
Chetan Parik is a customer support engineer from the Server Virtualization team at the Cisco Technical Assistance Center in Bangalore, India. He has seven years of total experience. He has worked on a wide range of Cisco data center products such as Cisco UCS and Cisco Nexus 1000V. He also has five years of experience on VMware virtualization products.
Remember to use the rating system to let Teclus and Chetan know if you have received an adequate response.
Because of the volume expected during this event, our experts might not be able to answer every question. Remember that you can continue the conversation in the Data Center community, under subcommunity Unified Computing, shortly after the event. This event lasts through May 9, 2014. Visit this forum often to view responses to your questions and the questions of other Cisco Support Community members.

Hi Jackson,
Yes its is possible. Connect the storage array to the fabric interconnects using two 10GB links per storage processor. Connect each SP to both fabric interconnects and configure the ports on the fabric interconnect as “Appliance” ports from UCSM
For more information on how to connect Netapp storage using other protocols like iSCSI or FCOE please check the url below.
http://www.cisco.com/c/en/us/products/collateral/servers-unified-computing/ucs-6100-series-fabric-interconnects/whitepaper_c11-702584.html
Regards
Teclus Dsouza

Monitoring Cisco UCS Manager via HP System Information Manager 7.1 (SIM)

I am working with a customer to configure HP System Information Manager 7.1 (SIM) to monitor their Cisco UCS Manager.
The customer is looking to monitor the following:
- CPU Utilization on manager, blades, servers, etc...
- Memory utilization
- Network utilization
- System inventory
Alerting is needed for the following:
- Hardware failures: memory, power supply, drive, etc...
- Predictive failures
- Alert messages
I have the list of all the MIBs provided by Cisco but an having the following issues while loading them into HP SIM.
While loading MIB "CISCO-UNIFIED-COMPUTING-TC-MIB" I get the following error message:
Line 128: Error defining object: expected a label, found reserved symbol {
Line in MIB: SYNTAX Gauge32 {
Guage32 is imported from SNMPv2-SMI MIB
To get past this error I found a version of the MIB that removes all the textual conventions that where causing errors. I have attached the fixed MIB file to this discussion. With the fixed version of the MIB installed in SIM everything compiles and installs except the following two MIBS. CISCO-UNIFIED-COMPUTING-NOTIFS-MIBCISCO-UNIFIED-COMPUTING-CONFORM-MIB Questions:
1. Is there any way to get the CISCO-UNIFIED-COMPUTING-TC-MIB MIB to install correctly into HP SIM?
2. Is my MIB load order setup correctly?
3. Has anyone had success getting HP SIM to monitor and alert for Cisco UCS manager?
MIB Load Order:
SNMPv2-SMI
SNMPv2-TC
SNMP-FRAMEWORK-MIB
RFC1213-MIB
IF-MIB
CISCO-SMI
CISCO-ST-TC
ENTITY-MIB
INET-ADDRESS-MIB
CISCO-UNIFIED-COMPUTING-MIB
CISCO-UNIFIED-COMPUTING-TC-MIB
CISCO-UNIFIED-COMPUTING-FAULT-MIB
CISCO-UNIFIED-COMPUTING-NOTIFS-MIB
CISCO-UNIFIED-COMPUTING-AAA-MIB
CISCO-UNIFIED-COMPUTING-ADAPTOR-MIB
CISCO-UNIFIED-COMPUTING-BIOS-MIB
CISCO-UNIFIED-COMPUTING-BMC-MIB
CISCO-UNIFIED-COMPUTING-CALLHOME-MIB
CISCO-UNIFIED-COMPUTING-CAPABILITY-MIB
CISCO-UNIFIED-COMPUTING-COMM-MIB
CISCO-UNIFIED-COMPUTING-COMPUTE-MIB
CISCO-UNIFIED-COMPUTING-CONFORM-MIB
CISCO-UNIFIED-COMPUTING-DCX-MIB
CISCO-UNIFIED-COMPUTING-DHCP-MIB
CISCO-UNIFIED-COMPUTING-DIAG-MIB
CISCO-UNIFIED-COMPUTING-DPSEC-MIB
CISCO-UNIFIED-COMPUTING-EPQOS-MIB
CISCO-UNIFIED-COMPUTING-EQUIPMENT-MIB
CISCO-UNIFIED-COMPUTING-ETHER-MIB
CISCO-UNIFIED-COMPUTING-EVENT-MIB
CISCO-UNIFIED-COMPUTING-EXTMGMT-MIB
CISCO-UNIFIED-COMPUTING-EXTVMM-MIB
CISCO-UNIFIED-COMPUTING-FABRIC-MIB
CISCO-UNIFIED-COMPUTING-FC-MIB
CISCO-UNIFIED-COMPUTING-FCPOOL-MIB
CISCO-UNIFIED-COMPUTING-FIRMWARE-MIB
CISCO-UNIFIED-COMPUTING-FLOWCTRL-MIB
CISCO-UNIFIED-COMPUTING-HOSTIMG-MIB
CISCO-UNIFIED-COMPUTING-IMGPROV-MIB
CISCO-UNIFIED-COMPUTING-IMGSEC-MIB
CISCO-UNIFIED-COMPUTING-IPPOOL-MIB
CISCO-UNIFIED-COMPUTING-IQNPOOL-MIB
CISCO-UNIFIED-COMPUTING-ISCSI-MIB
CISCO-UNIFIED-COMPUTING-LICENSE-MIB
CISCO-UNIFIED-COMPUTING-LLDP-MIB
CISCO-UNIFIED-COMPUTING-LSBOOT-MIB
CISCO-UNIFIED-COMPUTING-LSMAINT-MIB
CISCO-UNIFIED-COMPUTING-LS-MIB
CISCO-UNIFIED-COMPUTING-MACPOOL-MIB
CISCO-UNIFIED-COMPUTING-MAPPINGS-MIB
CISCO-UNIFIED-COMPUTING-MEMORY-MIB
CISCO-UNIFIED-COMPUTING-MGMT-MIB
CISCO-UNIFIED-COMPUTING-NETWORK-MIB
CISCO-UNIFIED-COMPUTING-NWCTRL-MIB
CISCO-UNIFIED-COMPUTING-ORG-MIB
CISCO-UNIFIED-COMPUTING-OS-MIB
CISCO-UNIFIED-COMPUTING-PCI-MIB
CISCO-UNIFIED-COMPUTING-PKI-MIB
CISCO-UNIFIED-COMPUTING-PORT-MIB
CISCO-UNIFIED-COMPUTING-POWER-MIB
CISCO-UNIFIED-COMPUTING-PROCESSOR-MIB
CISCO-UNIFIED-COMPUTING-PROC-MIB
CISCO-UNIFIED-COMPUTING-QOSCLASS-MIB
CISCO-UNIFIED-COMPUTING-SOL-MIB
CISCO-UNIFIED-COMPUTING-STATS-MIB
CISCO-UNIFIED-COMPUTING-STORAGE-MIB
CISCO-UNIFIED-COMPUTING-SW-MIB
CISCO-UNIFIED-COMPUTING-SYSDEBUG-MIB
CISCO-UNIFIED-COMPUTING-SYSFILE-MIB
CISCO-UNIFIED-COMPUTING-TOP-MIB
CISCO-UNIFIED-COMPUTING-TRIG-MIB
CISCO-UNIFIED-COMPUTING-UUIDPOOL-MIB
CISCO-UNIFIED-COMPUTING-VM-MIB
CISCO-UNIFIED-COMPUTING-VNIC-MIB
References:
ftp://ftp.cisco.com/pub/mibs/supportlists/ucs/ucs-manager-supportlist.html#_Toc303691433
http://www.hp.com/wwsolutions/misc/hpsim-helpfiles/simsnmp.pdf

Please post "debug ccsip messages".
Based on your debug you are getting "Cause No. 38 - network out of order."
You may want to bind SIP to an interface that the IP address is defined which Lync points to.
Chris

Cisco UCS with FEX system Component requirements

Hi ,
I am looking for quick suggestion for required components to have a VMware vSAN implemented using FEX system with maximum Fabric trough put. Ideally a configuration for without multipath using single Fabric switch and later on to be able to upgrade to Multipath fabric.
We are in great rush to provide POC concept and was looking if you folks can suggest me on this. If need be you want to talk to me more in detail I can available.
Appreciate your help here.
Who can I reach to able to pull up build and quote the configurations for implementing complete UCS solution.
Are these below components absolutely required ?
2 x Nexus 2232PP 10GE Fabric Extenders
The Cisco Nexus 2232PP 10GE provides 32 x 10Gb loss-less low latency Ethernet and Fiber Channel Over Ethernet (FCoE) Small Form-Factor Pluggable Plus (SFP+) server ports and eight 10 Gb Ethernet and FCoE SFP+ uplink ports in a compact 1 rack unit (1RU) form factor. They enable Seamless inclusion of UCS rack servers into UCS Manager domain with UCS blade servers when connected to the UCS 6200 Fabric Interconnects, providing converged LAN, SAN, and management connectivity.
We currently have these below servers :
2 x UCSC-BASE-M2-C460
        524 GB RAM
        No SSDs
        No VIC Card
        4 x SAS 1 TB Drives
        1 x L1 Intel 1 Gbps Ethernet Adapter
        1 x L2 10Gbps Ethernet Adapter
        1 x LSI Mega RAID SAS 9240-8i (no RAID 5 support, Need a card that supports RAID5)
1 x UCSC-C240-M3L
        132 GB RAM
        30 TB SAS HDD
        1 x VIC Card
        1 x Intel I350 Gigabit Ethernet Card with 4 ports
        1 x LSI Mega RAID SAS 9240-8i (no RAID 5 support, Need a card that supports RAID5)
1 x 5548UP Nexus Switch ( Will I be able to use this switch in place of Nexus 2232PP 10GE Fabric Extenders to achieve complete UCS solution)

Cisco UCS Manager 2.2 supports an option to connect the C-Series Rack-Mount Server directly to the Fabric Interconnects. You do not need the Fabric Extenders. This option enables Cisco UCS Manager to manage the C-Series Rack-Mount Servers using a single cable for both management traffic and data traffic.
If you need high performance low latency (storage, for VMware VSAN), direct connection to UCS Fabric Interconnect and/or N5k is recommended.

Cisco UCS RAID SAS 2008M-8i Mezzanine Card Performance?

I recently purchased a UCS C220 bundle, which includes a Cisco UCS RAID SAS 2008M-8i Mezzanine Card. I'm planning on deploying this as a standalone server running XenServer 6.2 in the near future.
I'm happy with the unit and testing is good, but for one aspect. The disk IO throughput on it seems to be far short of what I expected. I have a desktop PC with an Intel DZ77GA-70K motherboard as a lab spare, and the disk IO I can achieve from that server with the same disks exceeds what the C220 seems to be able to achieve, on a consistent repeatable basis.
The testing I am doing is based around two benchmarks:
1. A 400G file copy between the two machines, over the network back-to-back to note the maximum sustained throughput, and
2. A mix of runs with 'sysbench' to mix of tests of local IO, of sequential and random reads and writes, with the command: "sysbench --test=fileio --file-total-size=150G --file-test-mode=seqrd --init-rng=on --max-time=300 --max-requests=0 run"
For #1, I run a copy to/from an HP MicroServer Gen8, which has 12TB of disk space on it in a Linux RAID0 configuration (4x3TB Seagate drives). If I copy files to or from this HP Server to the Intel DZ77GA-70K I am able to easily saturate the 1G network, achieving a sustained 960MBit/s for an hour or more at a time. If I then take the exact same SATA disks from the DZ77GA-70K, connect these into the UCS box and do the exact same network copy with the exact same OS, I'm only about to get around ~400-500MBit/sec of sustained throughput.
For #2, the test results of which are entirely local to the C220, come in around 105-110 MByte/sec on a sequential read, which drops to around 2 MByte/sec on a random read or write test. No surprise of the enormous drop - because random reads/writes are a pretty tough IO load, but I would expect sequential reading should be much better. I can get consistent sysbench seqrd results from the MicroServer of around 300 MByte/sec, for example.
I can consistently replicate this with Redhat 6.5, as well as Gentoo (running the latest linux kernel) as well as from a Xen 6.2SP1 Hypervisor install on the C220 (tested from the Dom0 domain itself, as well as a Linux guest) all 64 bit. Jumbo frames are enabled end-to-end also, and CPU is not bottlenecking. Latest firmware is installed on all components. The ucs-cxxx-drivers.1.5.4a.iso image states that for the Redhat and Xen systems, that the required drivers are included in the OS, so I don't need to worry about installing them separately. Presumably the Gentoo system has even newer drivers again because it has a very new kernel, but alas the throughputs are the same on all of those systems.
I have tried with SATA as well as a SAS drive, and the test results are also practically the same. All disks in all servers are Seagate 6.0 Gb/s units, and none of the servers are swapping to disk at any stage.
I am happy with network IO - I can completely saturate the 1G ports easily, and I'm convinced that's not a part of the problem here.
What could cause this sort of performance? Storage card logs in CIMC don't indicate anything is wrong and none of the OS's are indicating issues of any sort, but it certainly does seem something isn't right in that I'm getting significantly superior performance from a desktop motherboard and the MicroServer, than an enterprise grade server, when testing with the exact same hard drives.
Questions:
- Is the 2008M-8i card considered a low-end RAID card or should I be getting reasonable throughput from it? I was anticipating performance at least as good as a desktop motherboard, but this doesn't seem to be the case. The RAID card as a component is more expensive than an entire MicroServer or Intel Motherboard so it should run much better, yes?
- What sort of performance should I expect out of this card on a single sequential read or write?
- Can this RAID card run drives just as JBOD's or do all disks have to be initialised in an array (even if just a RAID0 array with 1 disk)? It seems if they are added to the server they do not show up to any OS until they are initialised as part of an array, although I haven't delved into the BIOS settings of the card itself (only from CIMC so far).
- I recall seeing something about best practice of having two virtual drives on these cards, what is the impact in running more, given the card certainly allows more to be created (I currently have 4 while I am testing)
- I noticed on Cacti graphs while rebuilding a RAID1 array that the CPU ran hotter while the array was being rebuilt, and cooled down once the rebuild had completed, which indicates the rebuild was using up CPU on the host hardware. Should this not have been entirely transparent to the system if the RAID activity is offloaded to this card, or is an increase in CPU to be expected?
I'm very keen to find out others experiences of this card, what people have done to get good throughput out of it, or if I should go back to a whitebox server with an Intel board :-)
Thanks,
Reuben

Hello Reuben,
I reached out to colleagues who are more knowledge on this topic and here is their response.
- Is the 2008M-8i card considered a low-end RAID card or should I be getting reasonable throughput from it? I was anticipating performance at least as good as a desktop motherboard, but this doesn't seem to be the case. The RAID card as a component is more expensive than an entire MicroServer or Intel Motherboard so it should run much better, yes? -
2008M-8i card is a entry/value card with an expected performance which is better than a software RAID. This card doesn’t utilise memory in a standard RAID 0/1 configuration.
- What sort of performance should I expect out of this card on a single sequential read or write? - We should expect around 1GB sequential read (refer link http://www.cisco.com/c/en/us/solutions/collateral/data-center-virtualization/unified-computing/whitepaper_c11-728200.html)
- Can this RAID card run drives just as JBOD's or do all disks have to be initialised in an array (even if just a RAID0 array with 1 disk)? It seems if they are added to the server they do not show up to any OS until they are initialised as part of an array, although I haven't delved into the BIOS settings of the card itself (only from CIMC so far). -
This card can be used in JBOD mode. To enable JBOD mode you need to use MegaCLI commands. This option is not present as a default configuration.
Please note: Once we enable the JBOD mode – it cannot be reverted back to the default RAID mode setting.
- I recall seeing something about best practice of having two virtual drives on these cards, what is the impact in running more, given the card certainly allows more to be created (I currently have 4 while I am testing) -
This doesn’t apply for this card as it does not have any cache. Can you please point us to the document about best practice of having two virtual drives ?
- I noticed on Cacti graphs while rebuilding a RAID1 array that the CPU ran hotter while the array was being rebuilt, and cooled down once the rebuild had completed, which indicates the rebuild was using up CPU on the host hardware. Should this not have been entirely transparent to the system if the RAID activity is offloaded to this card, or is an increase in CPU to be expected? - Creating/Deleting/Modifying a RAID volume is a CPU independent operation.

Ask the Expert : Initial Set Up and LAN Connectivity for Cisco UCS Servers

Welcome to the Cisco Support Community Ask the Expert conversation. This is an opportunity to learn and ask questions about related to the initial setup of UCS C & B Series which include LAN connectivity from the UCS perspective with Cisco subject matter expert Kenny Perez.
In particularly, Kenny will cover topics such as: ESXi/Windows installations, RAID configurations (best practices for good performance and configuration), VLAN/Jumbo Frames configuration for B series and C series servers, Pools/Policies/Upgrades/Templates/Troubleshooting Tips for blade and rack servers, Fabric Interconnects configuration, general compatibility of Hardware/Software/drivers amongst other topics
Kenny Perez is a technical leader in Cisco Technical Assistance Center, where he works in Server Virtualization support team. His main job consists of supporting customers to implement and manage Cisco UCS B series and C series. He has background in computing, networking, and Vmware ESXi and has 3+ years of experience support UCS servers and is VCP certified.
Remember to use the rating system to let Kenny know if he has given you an adequate response.
This event lasts through October 10th, 2014. Visit this forum often to view responses to your questions and the questions of other Cisco Support Community members.

Hi,
Actually we have UCS 6248 fabric interconnect - first twelve ports are enabled and same in Cisco UCS Manager.
But when more port will be active by expansion module then UCSM can manage that too or need any other licence for UCSM too?

Ask the Expert: Cisco UCS Troubleshooting Boot from SAN with FC and iSCSI

Welcome to this Cisco Support Community Ask the Expert conversation. This is an opportunity to learn and ask questions about Cisco UCS Troubleshooting Boot from SAN with FC and iSCSI with Vishal Mehta and Manuel Velasco.
The current industry trend is to use SAN (FC/FCoE/iSCSI) for booting operating systems instead of using local storage.
Boot from SAN offers many benefits, including:
Server without local storage can run cooler and use the extra space for other components.
Redeployment of servers caused by hardware failures becomes easier with boot from SAN servers.
SAN storage allows the administrator to use storage more efficiently.
Boot from SAN offers reliability because the user can access the boot disk through multiple paths, which protects the disk from being a single point of failure.
Cisco UCS takes away much of the complexity with its service profiles and associated boot policies to make boot from SAN deployment an easy task.
Vishal Mehta is a customer support engineer for Cisco’s Data Center Server Virtualization TAC team based in San Jose, California. He has been working in the TAC for the past three years with a primary focus on data center technologies such as Cisco Nexus 5000, Cisco UCS, Cisco Nexus 1000v, and virtualization. He has presented at Cisco Live in Orlando 2013 and will present at Cisco Live Milan 2014 (BRKCOM-3003, BRKDCT-3444, and LABDCT-2333). He holds a master’s degree from Rutgers University in electrical and computer engineering and has CCIE certification (number 37139) in routing and switching and service provider.
Manuel Velasco is a customer support engineer for Cisco’s Data Center Server Virtualization TAC team based in San Jose, California. He has been working in the TAC for the past three years with a primary focus on data center technologies such as Cisco UCS, Cisco Nexus 1000v, and virtualization. Manuel holds a master’s degree in electrical engineering from California Polytechnic State University (Cal Poly) and VMware VCP and CCNA certifications.
Remember to use the rating system to let Vishal and Manuel know if you have received an adequate response.
Because of the volume expected during this event, our experts might not be able to answer every question. Remember that you can continue the conversation in the Data Center community, under subcommunity Unified Computing, shortly after the event. This event lasts through April 25, 2014. Visit this forum often to view responses to your questions and the questions of other Cisco Support Community members.

Hello Evan
Thank you for asking this question. Most common TAC cases that we have seen on Boot-from-SAN failures are due to misconfiguration.
So our methodology is to verify configuration and troubleshoot from server to storage switches to storage array.
Before diving into troubleshooting, make sure there is clear understanding of this topology. This is very vital with any troubleshooting scenario. Know what devices you have and how they are connected, how many paths are connected, Switch/NPV mode and so on.
Always try to troubleshoot one path at a time and verify that the setup is in complaint with the SW/HW interop matrix tested by Cisco.
Step 1: Check at server
a. make sure to have uniform firmware version across all components of UCS
b. Verify if VSAN is created and FC uplinks are configured correctly. VSANs/FCoE-vlan should be unique per fabric
c. Verify at service profile level for configuration of vHBAs - vHBA per Fabric should have unique VSAN number
Note down the WWPN of your vhba. This will be needed in step 2 for zoning on the SAN switch and step 3 for LUN masking on the storage array.
d. verify if Boot Policy of the service profile is configured to Boot From SAN - the Boot Order and its parameters such as Lun ID and WWN are extremely important
e. finally at UCS CLI - verify the flogi of vHBAs (for NPV mode, command is (from nxos) – show npv flogi-table)
Step 2: Check at Storage Switch
a. Verify the mode (by default UCS is in FC end-host mode, so storage switch has to be in NPIV mode; unless UCS is in FC Switch mode)
b. Verify the switch port connecting to UCS is UP as an F-Port and is configured for correct VSAN
c. Check if both the initiator (Server) and the target (Storage) are logged into the fabric switch (command for MDS/N5k - show flogi database vsan X)
d. Once confirmed that initiator and target devices are logged into the fabric, query the name server to see if they have registered themselves correctly. (command - show fcns database vsan X)
e. Most important configuration to check on Storage Switch is the zoning
Zoning is basically access control for our initiator to targets. Most common design is to configure one zone per initiator and target.
Zoning will require you to configure a zone, put that zone into your current zonset, then ACTIVATE it. (command - show zoneset active vsan X)
Step 3: Check at Storage Array
When the Storage array logs into the SAN fabric, it queries the name server to see which devices it can communicate.
LUN masking is crucial step on Storage Array which gives particular host (server) access to specific LUN
Assuming that both the storage and initiator have FLOGI’d into the fabric and the zoning is correct (as per Step 1 & 2)
Following needs to be verified at Storage Array level
a. Are the wwpn of the initiators (vhba of the hosts) visible on the storage array?
b. If above is yes then Is LUN Masking applied?
c. What LUN number is presented to the host - this is the number that we see in Lun ID on the 'Boot Order' of Step 1
Below document has details and troubleshooting outputs:
http://www.cisco.com/c/en/us/support/docs/servers-unified-computing/ucs-b-series-blade-servers/115764-ucs-san-tshoot-00.html
Hope this answers your question.
Thanks,
Vishal

Ethernet controller driver not installing after Windows Server 2008 installation on Cisco UCS blade

I am using MDT 2010 to deploy Windows Server 2008 to a Cisco UCS B200 blade. The OS and hotfixes install with no problem. The system reboots and attempts to connect to the deployment share to continue with the task sequence and I receive an error stating
"A connection to the deployment share could not be made. The following networking device did not have a driver installed. PCI\VEN_1137&DEV_0043&SUBSYS_00481137&REV_A2"
The driver in question is the driver for the Ethernet controller for the UCS blade. More specifically, it is the network driver for the Palo card in the UCS Chassis. When I Ignore or Abort the install, I am able to manually install the driver and restart the
task sequence. The build then completes successfully with no further issue. I just cannot seem to automate the driver install. Has anyone dealt with this type of issue?

Vik and stealthfield - thank you for your responses.
I was preparing a script to manually move the driver into the c:\windows\inf location when I had to stop to create a new task sequence for a different build. It was while building this task sequence that I found the soulution to the problem.
In the Post Install -> Inject Drivers step of the Task Sequence, I noticed that it was executing a command of type "Inject Drivers" with a selection profile of "All Drivers" and "Install only matching drivers
from the selection profile" selected.
I changed this to run a command of type "Run Command Line" and specified that it execute "%Scriptroot%\ZTIdrivers.wsf". I noticed that this was selected in one of my other builds that was working properly, so I thought I'd
give it a try. - IT WORKED.
So to sum up, the solution was to change the type of command being executed in the "Post Install -> Inject Drivers" step of the task sequence. After this change, the machine came up and continued the task sequence with no further issues.
Thanks again for the responses - They were helpful and I do appreciate it.

Cisco UCS Blade Deployment Without N5K or N7K ?

Dear all,
Is it possbile to deploy cisco UCS blade solution with only Blade Chassis, FI, FEX/IOM and using third party SAN and LAN switch ?
Thanks

Yes of course !
There are many installations with e.g. Brocade FC switches
The only caveat I see: are your uplink ethernet switches 1G or 10G; technically either one will work, however, with 1 G Northbound connectivity, some data flows are limited in bandwidth. see below

Cisco UCS network uplink on aggregation layer

Similar Messages

Maybe you are looking for