There are some tweaks and enhancements that I've already requested. One is to be able to make changes per device rather than as a global setting. That has to do with naming. It's minor.

The one area with room for improvement is probably administration. They added data insights to make a better user experience, but I'd like to see some improvements in the way the system is administered. 

We'd like the ability to do more templates at a global level to affect things across lots of places where we need to go and make adjustments. I'd like to see that become a little bit more unified. It's on the roadmap, so we're waiting to see what direction they take with it.

We have a real heterogeneous network with a lot of vendors and different types of equipment. Out-of-the-box, it doesn't tend to give us the things that we would like to see. Therefore, we have to raise certification requests.

I would like them to improve the self-certification, e.g., tools that allow us to certify products ourselves.

The event configuration piece could be overhauled. It is a little clunky and old compared to some things that I have used.

You need to plan integrations. That has been the biggest bug with SevOne so far. For the things that SevOne pulls directly, those are easy to understand, modify, and put into the database. For things that need to use the Universal Collector or xStats, you need to plan that stuff well in advance.

The user management features need to be improved. It would be nice if we had more granular control, or layers of control, out of the box. The way it is now, it is simple to configure, but if we want to add greater restrictions with more detail, it becomes quite difficult to do. 

Depending on the task, there are too many things to control. For example, there is no problem if I simply need to create a user role. However, if we have to create with LDAP then we need to manage many groups and different cases. This is something that I wish we could do differently.

One area that requires a little bit of improvement is the topology of visualization and being able to map out connections, end-to-end. It's able to do that, but it's not as impressive as we would like it to be. We would like to understand the different interface types and the connection points better, through the visualization. Heatmaps also need further development.

In addition, you can take a device and look at all the metrics that are being collected or enabled. But having a quick map view of the KPIs versus the alerting policies that we've built around a device, and being able to map that quicker and have a one-to-one correlation, would be useful.

We need to be thinking about streaming telemetry protocols. They already have the port for enhanced visualization, which they already have through Data Insight. I can't really think of anything else that needs improvement. It's meeting all the needs in those areas for now and the things they're claiming for the future are where we're hitting as well. There are some areas around multi-cloud or hybrid cloud solutions that we need to look at because we do have more of our workloads in the cloud so we need to consider how we can monitor the foreign stats in that regard. It's not something we've specifically looked at for SevOne at this point in time, but that would be something for us to consider.

The reporting of NMS is good, but it could be better. The challenges and deficiencies in the reporting are fixed with the Data Insight overlay. Generally speaking, the NMS reporting is excellent and it's fairly easy to use, but it can get complex as you get deeper into it.

In terms what could be improved, they need to integrate and get a better price. They can do cost analysis with Azure. They need to have a live cost analysis for the discounts, because if you have multiple thousands of VMs that you're doing, of course you're going to get a discount. Correct?

If you're only doing a few of them, you won't get a discount. That's the reason why they have to value the discount and coupons. The other con is that they need to be better with the accountability. In other words, the accounts or reports are not better than the others, compared with vRealize. The other thing is that you cannot write any kind of script in it to customize it to get other reports. So I'm shifting the gear into reports now.

There can be a problem of Microsoft versus Turbonomics. Because Microsoft won't allow the bigger clients to know what they're giving as a discount and they don't want Turbonomics to know what kind of discount I'm giving them. So there are pros and cons. Because these companies have a monopoly, they don't want the information of their biggest client to get out and say, "Okay, these are the coupons and these are the discounts I'm getting and let's see what Turbonomics can do."

High-frequency polling is data-intensive because you're pulling more. If SevOne could figure out a way to manage the impact of high-frequency polling on the system, that would be very popular.

The reporting is pretty straightforward but this is an area for improvement. The reports are easy to configure but they are a bit outdated in terms of appearance and visualization. SevOne has some alternatives where you can use Data Insight and it's easy to configure, yet outdated compared to other reporting mechanisms out there.

As we are moving to virtualization, it would be helpful if there was support for Kubernetes or microservices. If this added in the future then this might help us to better manage SevOne in a virtual environment.

One thing that comes to my mind is that while I was playing with the SevOne, when I started using it, I tried adding one of the BroadWorks application servers into SevOne. SevOne has all the templates for BroadWorks, but what happened was that it created thousands and thousands of objects from that one application server and we immediately ran out of license. That shut down SevOne. It was a huge pain for me to go into each object and disable and delete it from SevOne.

It would help, when new objects are discovered, if there were a way to categorize those objects and to pick the part of the object you need, rather than just discovering thousands of objects and adding them into the database.

SevOne NPM is good at data collection, but I think IBM needs to improve the solution's actionable insights. Many other vendors have machine learning or AI that pinpoint the potential problem for the customer or drill down to the root cause. I don't think SevOne has these capabilities at the moment. The cloud monitoring functions are also lackluster. Everyone talks about how good SevOne's cloud monitoring is, but I found it underwhelming. 

Telemetry is hot these days, and IBM can improve SevOne's support for telemetry correction. Reporting is another feature that could be better. It provides the bare minimum functionality, which is good enough for most engineers, but the management isn't advanced. The new portal provides a much lighter view and better visualization, but the management is not so good. 

You can use SevOne to monitor a mixed multi-vendor network, and it provides a baseline. It's a good platform, but we must rely on the implementation engineer who has the necessary knowledge to configure the monitoring policy for the customers. It would be better if they had some out-of-the-box policies that could help the customers.

SevOne Network Data Platform could improve its SD-WAN side. The system is still maturing. Cisco is always changing their product and new products are coming to market, they have an opportunity to focus and forge a good relationship with the SD-WAN product. They could build a strong product to provide services to SD-WAN.

I would like to see live maps as an added feature. Also, build modules on AI and EML to provide better data insights that would proactively tell us what we should be looking after.

NMS has several areas for improvement. It should be more user-friendly inside of NMS for some of the functionality in there. It's been getting better the last version or two, but there have been bugs in there whenever I've gone to new versions. 

There have been some features that were advertised that I would have that weren't actually there yet. They were kind of there, but even their tech support team didn't know how to use them because they were so new, and the documentation wasn't very thorough around those bleeding-edge features.

We previously have had discussions on some reporting enhancements. So, we raised a feature request, which was delivered from SevOne.

There are a lot of pain points. My main problem is that we don't have a high availability system. There are 20 peers. We're going to lose the end-of-life appliances that are old. If we lose a peer and it doesn't come back, we lose all that data. The reason we don't have high availability is because it's double the charge.

I wish there was some way that we could just get a snapshot of our system so that if one of our peers failed, we could go through the process and get it back to where it was. If we built another peer, and it took us four days to build another peer and get all the firewall rules and everything it would be nice when it came back if we had a snapshot that said, "Hey, peer two, that died." Then can we just slap all that data onto the new peer two and have all that historical data, as opposed to just importing it new, and it wouldn't have any data from the past. That's kind of a pie in the sky thing. But I would like some kind of backup system.

Keeping up to date with market trends, new vendors and with network vendors’ product support.

In terms of having a complete view of our network performance, I would rate it a nine out of 10. The reason for not giving it a 10 is that there is no packet capture associated with SevOne, but we do have other tools in place to do that.

In terms of stability, because of our move to VMs from physical appliances, some things have become a little unstable. It doesn't seem to be a SevOne issue, but we had to have a lot of calls with their technical support to figure out what's going on with it, but overall, it is pretty solid. 

The tool needs improvement in non-Cisco SD-WAN.

I'm not really sure if this was the software's fault or a server issue, but a couple of years back the disks were failing on our SevOne physical server every month and the server would go down. The secondary server took over from the primary until the disk issue was resolved. That was annoying.

Their virtualization solution is not compatible with our Kubernetes environment, which is one of the reasons we are ending our relationship with them. I didn't spend a lot of time evaluating with them why it was the case. It was simply not a roadmap item for them, so it was a pretty quick conversation.

Our version is quite old. In version 5.3.3.0, we see a lot of room for possible improvement. However, from SevOne support, we received confirmation that most of those expectations are met in version 5.4.x or higher.

Therefore, we have to think about upgrading to the later version as soon as possible.

SevOne should work with the graphs legend functionality. Now, you are able to put a part of the graph description as a customer description, but most of the original description stays unchanged. This means that sometimes the legend under the graph is unreadable (indicator names are sometimes not humanly readable). It would be nice to have a solution similar to the one in Cacti, where you can replace (or rather overwrite) an existing description with your own string.

We made assessments internally about the system, i.e., about what could be done better. There are a lot of points. 

With the administrative management of the appliance, if some object appears from SevOne because something changed in the network or whatever, then as an administrator you will not be aware. If you are using this object in a report, this object will disappear from the report and you will not be aware of it. So, if you have 1,000 reports, you cannot always check these reports everyday to see if objects are missing or information has disappeared. We don't have any information on alerts, saying that something is happening there and maybe we need to take action. If an object was replaced by another one, or if a link was replaced by another one, then the graph needs to be changed because it doesn't exist in the graph anymore. However, we don't have this information.

This is also the same in terms of the internal architecture that we put in place inside the system. We can tag our network device based on the firmware, some rules, or a manufacturer/vendor. But, it is not always clear when we add a bunch of devices that they will mark these categories. For example, we need to make sure that one device is from a specific vendor. We have to dig deep in each device to make sure they are really attached to the correct vendor. If they're not attached to the correct vendor, then information will not be pulled in the same way, and we might be losing information. So, small tweaks need to be made to the internal management, making it easier for me.

SDN networking is going from legacy towards object-oriented, so we don't have a choice. It is something that we are using and need. Unfortunately, our IT lead is not really into the SDN solution of ACI, as apparently it is missing some features in terms of deep analytics and monitoring of ACI hardware infrastructure as-is. In the future, we may use SDN infrastructure because I know they are releasing features every month.

I think that the downstream suppression could be improved. Suppression must all be done manually, but this improvement is on SevOne's roadmap, I believe.

Single device-based baseline monitoring works well in SevOne, but sometimes device independent base-lining in redundant network environments. Software upgrades can be tricky is not easy.

• Customization of fields
• The UI is a little dated, but it doesn’t cause any features to fall short of what they need to do.

In the area of server monitoring, it has some areas for improvement. For example, when you need to measure the CPU usage on Windows servers, it does not have the option to measure the total usage of CPU, only the measure of each core or CPU. Total CPU usage is something that's important to us.

The way they handle their data migration while upgrading the system needs to be simplified and standardized. I don't want to have to bring out professional services to help me when I upgrade to a new server. Every upgrade we've done -- and we've done them all -- have been highly disruptive and extraordinarily difficult. Upgrades of new systems should not be this complex. I have 125 appliances in my environment. The migration really needs to be performed in the most non-disruptive manner as possible. I can't afford anything less.

I've also found a horrible flaw in the migration process where the database is now using redundant disks. The operating system is not, so if I lose my operating system disk, I lose my system. That's a significant weakness because they basically rate controllers. Everything is rated a 5 except for the OS disk.

I'm not sure why this is the case. They have improved their reporting capabilities significantly over the years. I would like to see them continue, even to the point of maybe introducing a standalone reporting server. Either that or allow me to extract data into a data warehouse so that I can get more extensive reporting functions.

We're having that issue now where I am trying to extract data, so instead of keeping years' worth of data on the system, I keep three months of live data and then move the data offsite. It's been very difficult in the way they store data. It's inconsistent with the way other vendors do it, so it's been difficult.

We would like heat mapping and more detailed data, but in a general sense the initial mission was to get a more informed viewpoint of traffic in our network, and they knocked that out of the park in terms of those elements.

• Prevention
• Planning
• Service quality
• Problem analysis to help reconcile the part of the services provided

I can't think of anything as they have been very good to us, though it’s a little bit expensive. You could pay three times and get junk, or pay just once and get something good.

The method of searching for SIP and the way to create the groups.

I'd like to be able to stitch together multiple network elements or services for more of an end-to-end view of network performance.

I think that the upgrade process could be improved, which can be painful at times. SevOne knows it's an issue and they're working on it.

The product is perpetually evolving with a number of complimentary products in the pipeline. However, the basis of the product remains stable and any improvements are regarded as ‘enhancements’ and based upon industry market research at a given time and the facility to provide enhancement requests from individual customers.

The initial setup must be planned well to fit your environment.

I would say that incorporating other technologies, we have great performance data, but if it started doing intelligence and looked at the patterns or correlations, that would be great.

This is my favorite part of a review, so I am providing many ways to improve the product:

• Application session tracking
• Improve virtual appliance scalability
• Dashboard and report delivery
• Integration with service management (ServiceNow, Remedy) should be an out-of-the-box API

It needs a platform to add portals. Some of the low-level features and how they work could use some improvements.

A cause-and-effect kind of pattern would be helpful.

Enhanced baselining capability would be an improvement.

As I mentioned before, there are cheaper solutions available. Earlier, our management team only managed ICMP, like IP SLA. That's why, at that point in time, we decide to use Cisco IP SLA.

Well, I don't know if it's in development at SevOne, but some similar solutions offer end-to-end visibility, both regarding the server and also the network elements.

Other solutions also include the server and some additional layers, like an operating system or database, and in some cases, the application, too; their network elements are designed for management-level.

I don't know if SevOne plans on incorporating these features into their next release, but they should.

On a scale from one to ten, I would give SevOne a rating of nine.

There is no service mode setup in this monitoring tool if you want to snooze alerts for any specific amount of time, to account for any activity change or major incident. That is one of the drawbacks.

We have also faced some issues regarding SNMP traps.

Another difficulty is reporting. There is a limitation, per report, of 1,000 servers. Consequently, if your environment has 5,000 or 10,000 servers, you have to create separate reports, and each report will have 1,000 servers. That requires us to compromise with customers, that each report will only show 1,000 servers.

SNMP can't page exact disk utilization. So consider Unix servers. Unix servers' utilization is around 95%, and their reserve space is 5%. So SNMP will not page that extra 5%. it will only page the rest of the 90%. So from the OS end, it will see 95% usage, but in your tool you will see 90% usage because SNMP doesn't page that 5% of reserved space. That is one of the drawbacks that is not tool based, that would be considered as SNMP.

One other thing. Log monitoring is not possible from SevOne and that's why we are still using another monitoring tool. URL monitoring is also not possible.

Features like SMS support and a faster way to create objects in calculation editor would be good.

Continual UI improvements are needed to reduce number of clicks needed to produce a chart.

The GUI: both the dashboard/user view and the admin tool.

I think SevOne can improve by providing more details in their release notes regarding enhancements and changes. If the information can be organized in a way that highlights bigger changes it will help customers identify what areas will impact their operations most.

There's no need for improvement.

We use very few capabilities of this product, and from what we use I am very happy with in regards to what the tool offers.

But keep in mind that if you're using the the PLA version of SevOne, the hardware support runs out in five years, so if something were to happen to the hardware after this period, it would not be covered.

I'd like to see a configuration management feature that lets me customize analyses and data points important to me at any specific time.

User-friendly, multi-tenancy.