What needs improvement with Cisco Nexus?

Cisco's security features are in line with other OEMs, however, for data center security in terms of layer seven security, we prefer to have different OEMs rather than relying solely on Cisco.

In my company, we encountered some challenges since the network was not stable. The network is unstable, making it an area where improvements are required. If you do some kind of configuration on the Cisco Nexus and if something happens with the VLAN configuration, there isn't any mechanism that is visible to you. Adding additional visibility to VLAN that can easily be seen by experts will be one of the good features. We know that Cisco already has hardware-based visibility for the routers if you are actually planning to use that router as an SD-WAN. If similar kinds of features, if it was if it is actually included in Cisco Nexus for the VLAN part, then it will be good.

The product's high price is an area of concern where improvements are required. Cisco Nexus should lower its current prices.

As a device meant to lead the processes related to networking and troubleshooting, I feel both aspects can be described as areas with certain concerns where improvements are required.

The pricing must be improved.

The scalability can be improved.

Data center access is stable and high speed but it is not compatible with Cisco's server hardware devices. The solution can be technically improved and can have features like automation, better visibility, better functionality, and cheap pricing.

Talking about our environment and architecture, maybe it would be useful to have the whole configuration synchronized. Currently, the switches don't behave like the 4500 switches in terms of acting as one physical chassis. So the multi-chassis PortChannel feature is not applicable in this case. Maybe you could improve the timing of the settings. Currently, only the VPC configuration is synchronized, but other configurations (routes, access lists, etc.) must be manually configured on each chassis.

There is limited guidance in terms of the use and configuration of Cisco Nexus for specific use cases. The documentation is not comprehensive, and you have to do a lot of online research. I would like to have web interface configuration of the GUI in the next release. It would be good to be able to instruct via the GUI and then do the background commands. I would also like, in terms of tracking for HSRP features, active-passive links when using multiple active-passive links. It would be great to have data automation.

Areas for improvement would be the delivery timeline for the actual model, the length of which means we have to sacrifice a range of models because we can't wait six months for delivery.

Cisco Nexus if focused primarily on the data center. While it has been in the market for nearly ten years, there are still many areas in need of improvement. There may be many bugs which remain unresolved. It is important for Cisco to be aware of the various issues it encounters and come up with new releases. The frequency with which this occurs can vary. It may take months, be done quarterly or, sometimes, after half a year has elapsed. There is still much to be done to increase their stability, something which can depend. The solution is relatively expensive when compared with Cisco Catalyst.

The solution could be more user-friendly. It should have a more lenient cost. It could be cheaper in general. The installation process lasted too long, taking 16 to 24 hours for full completion. It should be faster.

The solution could be improved in general in all aspects. When it comes to technical assistance, you cannot speed up things. You can try to escalate, however, there's a process. You have not got them on speed dial. You need to wait. It would be ideal if there was better integration with various other platforms.

The solution is more expensive than other options on the market. HPE, for example, is a more affordable option. The renewal of support is quite expensive.

Cisco Nexus could also have better documentation. Moreover, it would be nice if the solution offered multiple types of port support. It does not have, for example, a single switch for both internet and fiber ports.

I'm not in a good position to comment on what might be lacking. I just use the Nexus switches, and I'm very happy to use the machines. A lot of technologies and features are present on Cisco Nexus and I haven't had a chance to dive into all of them just yet. We had some issues, strange issues with our firewall from Cisco, however, we resolved them with Cisco. They were very, very strange issues indeed. However, they seem to originate only on our site, and not from Nexus. From my point of view, I've never seen in 20 years, a single breakdown. They should work to make the pricing more reasonable for the local market.

The implementation process could be easier, which is something that should be improved.

The software is not as mature as it could be and needs some integration improvements with other orchestrators.

The cost of the support can be improved. We had critical operations, and we needed 24/7 support for 365 days, which was quite expensive. We had to go for a very costly support contract, which was really a concern. The availability of spare parts, especially in a remote location such as Egypt, can also be improved.

One of the biggest challenges, which I see is that there's a constant evolution in the product. For example, our configuration is based on what is known as traditional data center implementation. Today there is the ACI deployment and to implement, to migrate from one technology to another, that's challenging both from a configuration perspective and also from a cost perspective. We have had a problem in an order of a batch of optics, which is practically failing. So that was about the average of optics, which caused us quite some problem.

The price could be better.

There is an ongoing problem with the limitation of the TCAM table, which is that it doesn't have enough memory to allow you to be really granular with your policy. Without enough memory, it requires manual manipulation if you exceed or get near to the TCAM limits. They have improved it, but in the early days, it took down some companies.

They could improve on having different technologies between product models. The management dashboard could be better if it could handle the Catalyst and the Nexus at the same time. If Cisco could find management connectivity between the DMA center and the ACI connectivity this would be helpful.

I feel that this solution should be more flexible and scalable.

I am looking for a GUI that goes alongside them and more SD-WAN built to their core switches. I'm looking at changing our firewall estate, and at the same time, looking at integrating our firewalls with whatever our core switch solution is going to be. I can't necessarily do that with the costings of a Cisco organization, as they're too expensive.

The licensing is very complicated. They should work to simplify it. Every feature you want to use you have to pay for separately as an extra cost. The features should be bundled together. It sometimes causes us to have clashes with dealers. We don't find that there's any difference between the Nexus and Catalyst solutions. It seems that just the marketing is somewhat different, even though they seem to be the same technology.

I don't really know any improvements that we would need right now as a company. I would say it would be nice to support more phone models in general. For us that doesn't matter as we really stick with one type of phone. There hasn't been a phone refresh — I think — in maybe five years, but I think that's kind of how Cisco rolls. They do have a 10 year run on phones. Better monitoring would be a big thing to have. The RTMT (Real-time Monitoring Tool) is good, but I also know they're moving to a web-based solution so certain updates to current products won't be on the way. Being able to drill down and have better adaptivity going forward would be nice. But I think that concentrating on the web-based solution is their plan and the way they're going. I'm looking forward to seeing how that works out.

They should make sure that the back address auto is baselined. I think it might even be baselined, so it might be that one of our team members had messed that up, but it just wasn't a very straightforward command. You should have multi-context, multi-port channel enabled on the underlay of the Firepower. The way the Firepower firewalls work is they have a management plane, and then you make the virtual ASAs on the Firepower, where you're assigning multiple interfaces or core channels on there. have fewer devices and fewer physical interfaces, yet retain a more extensive logical setup.

I would like to see more automation and for it to be easier to use.

In the next release of the solution, I would like to see varying integration between VRF (Virtual Routing and Forwarding) instances so we can integrate BGB into different VRFs in a more organized way. We have some issues with tunnels and VRFs.

I would like to see the usability improved by simplifying the user interface. For example, it would be nice to have a simple way to find endpoints and get information about them. It would be great if they could make the interface a little more user-friendly, but not trade power off for simplicity.

In the next release of this solution, I would like to see a focus on backplane deployment, so you can stack more.

Nothing is perfect.

Our only complaint is about the licensing because it can always be a little more cost-effective.

There is some room for improvement when it comes to the frequency on the network.

Overall, I wish it was a more intuitive OS.

I would like to see some development into the SAE world where you can virtualize. That would be the only thing I would like to see because I need to start virtualizing my infrastructure gear at some of the other data centers that I have.

In terms of improvement, the stacking between the two switches in the higher availability model should be similar to Catalyst. It's better to have one interface for configuration. I hope to see more features for high-availability configuration such as a VSS configuration on the Catalyst series. To have one management console for both chassis. I also hope to see the dependency on the management interface for high-availability lowered. What needs improvement is the price on the DNA subscription. It's obligatory for all the switches and you should have the option to remove it or keep it as optional. Cisco has the option to remove it now. We have agreed that in the future, they will offer a permanent license for it.

The VLAN is a little more hand driven now. I don't know exactly why the split screen does that. If we could get some more automation integrated into it, it would make it easier.

I would like to see network function virtualization with no hardware. I would like to see the integration of the products into something where it's seamless where an engineer never touches the switch again, never does CLI and you move to an application based network organization.

I would like to see better collaboration with other, low-end devices.

One improvement needed is support for Multi-CAD scale that we were concerned about. We're not hitting any limits at this point. There were concerns about the amount of server capacity that was going to be available. We like to see things that are already there as opposed to being told where they will be.

The price is a con.

I get more of the frustration out of Juniper equipment than I do from Cisco products. Cisco Nexus is more versatile for us in comparison to Juniper. The routing is easier with Cisco.

I would like to have the feature where you can install something like operf on it so you can do troubleshooting without having to configure operf on servers. We had issues with Cisco Nexus during upgrades. When we first got the system, we were having some issues. The whole system crashed.

They should make it easier to update the code on it. Upgrading should also be easier.

Cisco Nexus is very new. We're still on the learning curve for the 9000.

Cisco Nexus still needs more stability. Not that they're unstable. FlexPod in particular. My team doesn't work with that but it's owned by the server infrastructure team. It's in the UTS. They're pretty stable but every once in a while, it's not.

I would like to see more granularity.

The additional features I would like to see included in the next release is the ability to integrate routing and switching features.

I would like to see Duo directly integrated with the Nexus hardware, instead of it being through the proxies, like how Duo works today. The way it works now really irritates me. There is a lot of depth in the Nexus class, and it would help to make this more accessible to the users. We, for instance, are using them no differently than we would a Catalyst switch with VPCs. With that in mind, some of the more in-depth features would be nice to have, so having easier access to them would be beneficial.

We have encountered some software bugs.

The technical support for this solution needs to be improved.

Upgrade and promote catalyst switches to the Nexus, it's a lot faster.

I would like to see better monitoring actions supported in the next release of this solution. An overview of what is going on with the system would be helpful.

We really lag on the AC infrastructure where the configuration and the changes make a difference in terms of troubleshooting. The product improves us when we see an expert and L1/L2 engineers for support. We have a dependency there. A simple configuration makes a big difference. It can create more chaos inside the network. We need to make sure when we make changes in one platform and it impacts other platforms, that the technical problems don't reach the end users. There should be better certifications. More training should be provided before we get into this product.

There are still bugs in the system that need to be fixed. Technical support could be better.

Enhancing the software-defined level of this solution would empower it.

I would like to see a central management solution in order to have all of the equipment in one place. Also, more documentation and more design examples would help us to reach other clients.

This solution is only scalable to a degree. This is a very expensive product. Areas in which the solution have room for improvement are the interface, management, and reporting. You still have to go deep into the CLI to find issues.

We have uncovered things about some of the features where there is room for improvement. * There are limitations on some of the lower level 9K solutions where you can't do the same things that you would normally be able to do, like for instance the number of static maps that you can configure on it. * There are limitations between different product IDs that I do not really understand, and I don't think that there is a really good reason for.

The 9K was developed to support ACI, the software-defined data center technology. For this reason, there is no feature parity between the 9K and the 5K or 7K, which means that it is harder to position the 9K to future proof them. If a customer wants to leverage their investment for ACI in the future then it is difficult. I know that Cisco has added some Fibre Channel over Ethernet capabilities to the 9K line, but there are some other features that it does not have capabilities for. For example, virtual device context is not supported. It would be really nice to see some capabilities like that added to the 9K line so that we can position them to future-proof our customers. I understand why it is that they don't have some of the features from the older Nexus models, but we get into some scenarios where the customers need those features, and they have to go with a 7K or a 5K. Ideally, I would prefer to position a 9K if I could, to future-proof them and lead them along that path to ACI, eventually.

There is always room for improvement.

My only complaint is that not all of the ports on it are enabled from the start. We pay for twenty-six ports, and we would have to license the other ports if we want to use them. This can get expensive as you expand more.

Any problems we have with the product stems from the difficulty of connecting with other non-Cisco products. There are other products on the market which we would like to have integrated. That's when we have a real solution.

I would like to see more on-device programmability, as it seems to be lacking in this platform. The EEM has limited capabilities compared to the Catalyst 9300 series, so this can be improved. I would like to see better support for the newer front end tools, such as the DNS center.

I would like to see improved ISSU.

Currently, the platform is heavy. I would like to see a lighter platform.

Some of the DHCP features need improvement.

The only real feature that we are hoping would evolve at some point is multi-contexts to get into multi-tenancy type environments. We are very interested in that with the 7000s. But we didn't like all the other robustness that 7000s had that we had to pay extra for. 9000s fit our needs a little bit better, but multi-contexts would be a really good feature for organizations like us, so we can separate different tenants logically and in a more secure fashion.

There is not much that needs to be improved at this stage. If it were possible to gain even more throughput and more port diversity, that can always be a benefit.

For a very small subset, Cisco Nexus caused port flapping within the applications, which was weird. Call it an outage. It was one specific application, i.e. because of the way the application was routed through the switches which couldn't be handled that way. We had the good side because 90% of it works flawlessly. We didn't have to repeat IP everything. We didn't have to go through and take everything down. We just migrated it. But that one small subset went haywire.

We would like to see this solution support routing. One of the problems we have is that there is no confirmation for when you try to test a system command. It just accepts it and does not give you a response.

We would like to see OpenConfig covering most of the industry standards. The YANG models, for example, and also support for OpenFlow. The areas that need improvement are MACsec, OpenConfig, and OpenFlow.

I would love to see Active/Active FEX connectivity, which is enhanced vPC, where we could also do a vPC from a server in that type of configuration. It's still a limitation that's been carried over from the 7K that I'm surprised is still absent from the 9K. Including this would give us the highest level of redundancy without risking having orphaned ports, which we do have at times.

I really wish that Nexus and iOS commands were more similar than they are different. Standardization across iOS and Nexus operating systems would be an improvement.

They need to offer fibre switches in the 9K series. I would like to have the ability to identify ports by flashing port lights, so I can tell the user "We need to look at port four", and then I can flash the light remotely using the command line.

This solution costs a lot.

The code quality for this solution has gotten worse and needs improvement.

The only issue that I have with this product is finding help on different commands when I'm trying to make a change. Or, if I'm troubleshooting an issue, finding what commands I need to do certain things. I know what I need to do, but I'm not sure of the exact command.

Being this is the first usage, we don't have any expectations yet.

The product could be improved with a friendlier GUI.

The solution has room for improvement in terms of the learning curve.

In terms of the context of this deployment, Cisco Nexus did everything it needed to do. I'd like to see this be wrapped into ACI. I'll be more comfortable when it's had more deployments. There's more knowledge out there on how to configure it. The Cisco Nexus is a big change. I'm not currently looking for new features. I am looking to other customers with good experiences deploying in ACI mode.

The only major improvement required would be stability. With some of the products we had before, we had a little more downtime than we would like. We had to spend more time either dealing with parts that needed to be replaced, or issues that we had in configurations that we needed to upgrade. The Cisco Nexus is a lot more stable and doesn't have all of the bugs when it has to do with upgrades. Some of the fiber optic capacity increases will be good because we're already looking at 40 GB and 100 GB at a reasonable price. Fiber capacity is going to be something we're starting to look on our roadmap, how do we increase what we have available. We use Cisco Nexus and haven't had any problems. We've been happy with it.

In the Nexus 3500, you can't do per-port Jumbo frames, so you have to enable it as a global configuration. This is a feature that should be added.

Cisco Nexus is a switch. It's fairly straightforward in what it does and what it can offer. Other than the differences between the command line or vs. Catalyst, it's just getting used to it. I don't have any negative issues. Cisco Nexus is expensive.

The additional features I would like to see is X-LAN and ACI. More ACI or SDN features.

In future releases, it would be good to have more management on those devices. Cisco needs to provide more training.

In the next release of this solution, Cisco should keep trending on ease-of-use, with more human or regular language and commands. Maybe they could add some machine learning and AI integration. Ease-of-use and making things more human-readable is most important.

There are too many different model numbers and too much complexity. The Nexus 9K, for example, came from a family of twenty offerings.