DX Spectrum Reviews

• Network Configuration Manager
• MPLS & VPN Manager
• Southbound Gateway
• Event rules
• Event procedures

Using event rules and procedures, we improved our event management process. CA Spectrum is very useful for event handling and offers a robust manager of managers console.

Like SpectroSERVER's fault tolerance (FT) functionality, the OneClick console should have a native HA or FT functionality. Actually, the only way to get this is by using a load balancer.

Spectroserver is the core component in CA Spectrum arquitecture that polls devices in to network, you can use multiple servers to build some kind of "cluster" to provide continued device polling when some server fail. Oneclick console ( web server) is front end of CA Spectrum it do not have an option to provide continued user access when a oneclick server fail ( High Availability). You can use a Load Balancer but is very expensive.

I have been using it for seven years.

No deployment issues encountered.

SpectroSERVER and reporting databases can grow quickly if you don't keep track of trap storms; this can cause SpectroSERVER to crash and performance issues with reporting.

No scalability issues encountered.

Customer service is very good (10/10).

The North America support team is excellent (10/10). The LATAM support team have some things to improve (6/10).

We did not previously use a different solution.

Initial setup was complex because deep knowledge about tool architecture and proprietary software is required. In some cases, if you make a mistake during installation or initial setup, the better solution is to reinstall it or reinitialize a DB.

We implemented it through a vendor team. Their level of expertise is high (9/10).

Licensing usually is expensive but cost efficient.

Before choosing this product, we evaluated Tivoli monitoring.

I recommend you use CA Spectrum mainly to monitor network infrastructure. Use this tool as a manager of managers. Event monitoring is a very powerful functionality. If you want to monitor servers, app and databases, use CA UIM probes instead of the CA SystemEDGE agent.

• Works excellent for heterogeneous networks (lots of different vendors)
• Error correlation
• Root-cause analysis
• Service intelligence
• Lots of add-on products for Spectrum, e.g., VMware integration
• Local extensions/developments for special needs, e.g., new vendor xyz

With the object-oriented design of Spectrum, from its very first version, we had very important issues, such as fault isolation and root-cause analysis implemented by default!! Which means that we had through all these years very precise fault indications on where to look for the error and were not overwhelmed by long lists of alarm, which were only effects of the real fault.

So we then concentrated on building a virtual picture of the complete real IT infrastructure inside Spectrum. This includes today a very wide range of devices, which are modelled inside Spectrum and interconnected in the correct hierarchy as in the real environment. And, as the big buzzword today is IoT, this means for us as a service provider to our university users to deal with MoT = Monitoring of Things!!

The “new” things include devices such as coffee machines, fridges, temperature sensors, worktime badge systems, door opening scanners, house automation systems (for light, sun blinds, climate, etc.), smoke and fire detectors, etc. This list increases every year with new “things” popping up on our network radar!

Of course, the real IT devices like routers, switches, firewalls, servers, storage devices, WiFi access points, IP phones, and, importantly, printers, form the base of the big “virtual” Spectrum picture.

And last but not least, Spectrum by design can be expanded in its function by a series of add-ons, such as VMware integration, which reads out automatically the virtual servers in the vCenter and models all VM’s inside Spectrum in the corresponding physical servers. Another example of an add-on is the integration of a performance probe with scenarios that run periodically (i.e., log in to a webmail portal, check for the existance of a message xyz, lookup the calendar, etc.) and map those scenarios to objects within Spectrum, which can be handled like devices, so you can put thresholds on them and get alarmed when violated!

Thanks to Spectrum, we are able to monitor (and send out alarms for) our whole IT infrastructure 7/24 without having more than a minimal number of technical persons.

• Reporting
• Full web-based topology view
• Path view

I have used it since fall 1990. We’re a simple - but extremely long-term customer.

I only encountered stability issues in very rare situations; mainly when we used our own extensions for private management modules.

I have not encountered any scalability issues so far!

Technical support is excellent - 9.5 on a scale 0-10 - but this also includes support from the local Spectrum partner here in Switzerland.

I did not previously use a different solution. At the time (fall 1990), there were only very few products available, where just two or three of them had a graphical visualization.

Initial setup was completely straightforward, as the product already included a discovery and an auto-connect function.

Start small and include as much of your IT infrastructure as possible, so that at the end you have a picture of the “big” total.

It is our network infrastructure monitoring tool.

The product’s stability and robustness are valuable. It is capable of efficient fault management irrespective of the size of the network.

The product should provide performance management features.

I have been using the solution for around 20 years.

I rate the tool’s stability a ten out of ten.

I rate the tool’s scalability a ten out of ten. In my region, five to ten customers are using the solution. We are a small market in Turkey.

The technical support is very mature. The team provides good implementation support.

The initial setup was straightforward. The time taken for deployment depends on the project and architecture. Simple projects take only a couple of days.

The product’s price depends on the deal we make with the vendor.

The solution releases new updates every quarter. It is easy to implement. It is the number one product in the market for complex and difficult-to-manage networks. It covers all our needs. The solution’s capabilities are far better than its competitors. Overall, I rate the tool a nine out of ten.

The primary use case is eyes on glass, operations, responding to tickets, alerts, and incidents tickets. Performance wise it is good. We have a lot of different monitoring tools in the background, so orchestration has been a little bit of a challenge, but I think that is everybody right now. Overall it is a good tool.

• The ability to quickly identify and isolate incidents.
• Automate ticketing

With the service manager today, we are moving the service now into the future.

A lot of the filtering and orchestration of the background takes a lot of manual work. There are a lot of processes involved and it is hard to keep it up-to-date with a lot of moving parts. Making it more dynamic could be better. 

Stability has been good. We really have not had any issues.

Scalability has been good. 

I have not personally used the technical support.

I was not involved in the initial setup.

It is good, but there is a lot of growth opportunity in the space. I think CA and a lot of other companies are moving in the direction of orchestrating cross-platforms. Right now, the way we are using Spectrum is really only covering infrastructure and infrastructures out of the house. I think the future with APM and other tools, being able to see more holistically what's happening, is very critical. 

The most important thing is implementation, how it is implemented to enable the operations. In my current role, the operation's visibility and usability are the most important thing. So, making sure it is implemented in a way that it meets the requirement of the operation space. 

Spectrum is our correlation engine and an integration point of external monitoring tools. Also, we send UIM traffic to it, then we forward it onto SOI.

• The correlation
• The integration point

The biggest issue is our integration right now between UIM and Spectrum is lacking. It does not allow us to pass all the details. There is a new integration, which has the new version of UIM. I hope it clears up everything.

Spectrum has been very stable.

I do not think we had any scalability issues with Spectrum. We have one of the largest environments, and it is handling it just fine. 

For Spectrum, it is good. For other things, we have had some issues. We had a six day outage once on SOI, but we worked with CA to try to improve that. 

The negative is they have a tendency to ask us to do things that basically we have already done, like balance a service or disable all the policies or something like that. You can tell, they are just walking down a script. This not going to help; logic says this is not going to help anything. 

It would be helpful if they would try to look at the situation and decide this is something that really needs to happen or not. Instead of just doing it to say they have done it.

I was not involved in the initial setup. However, the biggest issue is setting up some of the custom traps is a little complicated. So, you have to have someone that really knows the tool. It could be a little more intuitive overall.

A valuable feature is the topology view. You can view the topology. Other valuable features are the host configuration, the NCM feature, the policy manager that it has and the scalability. That's one of the biggest ones.

When I started working, we only had 35,000 devices in there. Because of the scalability and all that, we have over 100,000 in Spectrum right now. All of them are SNMP managed.

We have a DR site. We are testing the latest release right now. In our production, we have 9.3.3 and it's running good. With our requirements as they are expanding, for example, one of the issues that we have is the global collection, the real-time global collections. For example, let’s say, just the southwest region has 15,000 devices, and that group wants to have interface global collection; that's going to be over 300,000-400,000 interfaces, and they wanted real-time interfaces. That really bogs down the database, so we are looking at 10.1. In there, the scalability is going to improve a lot with that.

There are certain attributes that we would like to be able to send northbound to different applications or different API calls that we can use. That is one area for improvement.

Also, I would like to see a unified poller, which they have already talked about. If I have Performance Center in Spectrum, I don't need to have separate servers for those same applications. I could have one unified poller, which is getting the traps and also polling information.

I have been using it for two years.

It is a very stable solution. Once you launch it correctly, you barely have to manage it.

The level-one support is the same as the support you get everywhere with every vendor, but the layer two and above, they are very good.

I am not sure how they knew they needed to invest in a new solution, but when I got in, Spectrum was on the rise of being implemented. Before this, we had NetCool.

When I select a vendor, first, I look at the product; how viable it is. Second, I look at the support that I'm going to get from the vendor. Those are the two major things that I look for.

Initial setup complexity depends on how big your environment is. If you're doing a distributed setup, then you need little bit of architecture help on how it's going to be built, but on a scale from one to ten, with one being easy and ten being hard, initial setup is between six and seven.

My company, they looked at vendors before Spectrum, but now Spectrum is the primary vendor for fault management.

My rating reflects its scalability and reliability. There are some minor issues, but that can be resolved.

It's a very mature product. It has a lot of features. It has fault root-cause analysis, so you can see a fault and know that those 10 or 50 additional faults are all related to that one. You don't get 50 messages. You just get one root cause.

It’s given us a faster response time. It’s been able to pick up things before end users call about them. It provides us with a lot of extra information about how the network health is.

In the latest release, they've added 64-bit performance. They've added several enhancements to integrate with other products. It's compatible with other products, and it gives us more information that way.

I would like to see more integrations with other products.

It’s been very stable for us. We don't experience many problems.

It's very scalable.

Support has been doing a good job. There have been no problems or issues.

Definitely consider Spectrum. It's one of the top solutions for its market space. It's not very hard to set up. It does a good job.

When selecting a vendor, we want something that’s been around awhile. It should be a mature stable product.

We monitor our network devices with it, therefore it performs well.

The flexibility. It is so customizable. You can about make it do anything you want it to do, if you can figure out how to make it do it. There are things you can do in it to customize it to your environment and make it meet your business needs. I do not know if I have seen other products in this space that are as flexible.

We see events that we would not typically see. Thus, it helps us find things that we may not have known were wrong. It creates transparency and makes everything work by being more transparent.

CA is going in the right direction. What we want is more integration between the tools and it looks like they are doing more of that. That is probably the big thing. Better integration between the tools, being able to take all the data from the tools and having it available in one place, such as generate dashboards, analytical data, etc. 

I would like to see CA move Spectrum, Performance Manager, and some of their other tools moved to a SaaS based model. You see a lot of other manufacturers which are moving toward SaaS based models.

The stability is good. We have not issues in the two years that we have been using it. It is one of CA's more stable products.

The scalability is good. We have not scaled it a lot. We are getting ready to add another SpectroSERVER to it for redundancy. But from a scale perspective, it has worked out okay for us.

Technical support is so-so. It depends on who you get. Sometimes you get the runaround. Sometimes, I am not sure that the people that pick up cases know the product well.

You open a case specifically for Spectrum support, but depending on who you get, some of the guys are really good or knowledgeable. Typically what you find out is those are kind of the higher level people. Then, sometimes you get people that you do not get really good answers from, who make you feel like you know more about the product than they do. So, it is hit or miss. 

Sometimes, you have cases that may stay open for weeks. These are not complex cases. These are just questions about how to do something or if you can do a certain thing. It seems like it takes a long time to get an answer. It just depends.

The communities are probably a bigger help than actual technical support because you have a lot of other people that can provide feedback. Therefore, it is not terrible.

It was pretty straightforward. We actually had the CA Services guys come in and help us set it up. 

We were involved in the the set up and everything, but CA Services did the heavy lifting. They were the experts, so we did not have to have a ton of knowledge. So, the setup for us was pretty easy because we used CA Services.

We had HPE OpenView, SolarWinds, and OPNET tools. We chose CA because we can consolidate all of those tools down. This is why the integration is important. It has allowed us to eliminate a whole bunch of other tools, so we were able to consolidate, and that was a good thing for us.

It is a really good product, though it is very complicated. You need to make sure that you have the staffing, because of the complexity of the product. You will need staff to be able to use it to its fullest extent. 

Most important criteria when selecting a vendor: 

The biggest criteria are:

• Functions
• Features
• Cost
• The reputation of the vendor.

There are a lot of little vendors out there that are in a lot of these spaces, but you have never heard of them. Therefore, it is good to have a large company that you know and you trust. We have a good account team and good pre-sales people. They have really help us out a lot.