DX Spectrum Reviews

We use Spectrum to monitor our infrastructure, servers, network, routers, switches, to make sure that all of our services and infrastructure components are all up and operational all the time. And if they're not, we're responsible for alerting the people who are responsible for the different components and parts. It's a pretty important part of our operational M.O. as we move forward, making sure stuff is up and running.

The performance has been fantastic.

Since I put this solution in, over the last five years, we have actually reduced the number of alerts that go out and wake people up in the middle of the night by 70%.

The stability of the product is key. It never breaks, really. The stability and reliability of the Spectrum product have been top-notch.

The only reason you're not going to know that your stuff is up and running is because you set it up wrong, not because it's not doing its job.

Better integration with all the other products in the Agile suite of tools; anything they could do to make that less complex, would be great. They've made it less complex, it's getting better as time goes. It's just such a long road because different products were originally different companies that have been hooked together, acquired, and then assembled as their overall solution. And there's good and bad in that.

The good is they have all these capabilities in a one-stop shop. That's awesome. But, sometimes those were competing products that are now having to work together, and they have to come in and force them together. It adds to the complexity.

If nobody else has said documentation, let's go there. I understand, you can either write about it or you can do it. And most of us would rather they do it, but now that they've done it, those of us that didn't do it, we need to go and find: "Where did they write about this to tell us how to do it?" That's always lacking.

Some of the products are better than others at documentation. Some are them are very good and some of them are not as good. It's a little disheartening when you go in to find whatever it is procedure you need to do and it's just not there. Or it's wrong. That's my favorite.

It scales way beyond what we need. I can actually do everything I need to do on one server, even though I have multiple for redundancy. I'm not going to say that I've not had problems with failover and all, because I haven't had to do that; other than me forcing it to do it and switch around. It hasn't actually broken on its own so it hasn't needed to activate its redundancy.

The support folks, that's another aspect of it; partnering with CA to make sure that our stuff is up and running. And if there is a problem with it, they have some really decent people that will get on and help you come up with solutions.

Tech support is great. I usually fill out the surveys when they help me. I haven't had many really bad experiences with support. The people I've worked with, if it's a critical issue, they are right on it. If it's more of a generic thing like, "How do I do that?" they get back to you within a reasonable amount of time with the answer.

I did the architecture and implementation, soup to nuts; made the Kool-aid, served the Kool-aid, drank the Kool-aid, every day.

The nature of what we do in this space is complex anyway. There's only so much you can simplify at this level of complexity. And as much as it is complex, they've made it a simple as they can.

I think the short list for Spectrum was Solar Winds, Nagios, Spectrum, of course, and BMC.

You look around and see there are a lot of home-grown things, and multiple different solutions from different vendors that just don't work together, or they can't work together for whatever reason.

One thing I like about the one-stop shop with CA is we don't have to have staff to be experts in 84 different things. If I need these two things to work together, if nothing else, I can call support and say, "I tried this and it's not working. Why is this not working?" And they'll get on the phone and WebEx and look and eventually you will get it working.

When it comes to my most important criteria for selecting a vendor, ease of use is always one of the top ones: How many people in wagons and trucks, and servers, how many things do I have to have to run this tool, this product? That is the key factor in my evaluation.

Of course, price, obviously, that goes without saying. The cost of some of these products are a little cost prohibitive, which is unfortunate. We have to find lower cost solutions for some of the things that we have to pull in. 

For the most part, CA has capabilities in pretty much everything that you would need. You just have to figure out where are you going to spend the most money for what you have to actually get done. You can do all these things but, which ones do you have to do? And that's hard to choose sometimes because you want to do them all.

Mostly SNMP trap management.

I am fairly new to the product end of the solution. We have not utilized it to its full capabilities yet. I think we are still trying to figure out how to get the most out of it. 

The network mapping functionality.

I would like to see better scalability, and maybe, a more intuitive user interface. 

It seems fairly stable. We have had some issues with the spec server and the high availability replication of data. It fails over well, but it does not come back very well. We get duplicated events. 

It seems okay, but not great. 

Good, if you get the right people. 

I was not involved in the initial setup.

I have looked a little into open source solutions.

The product is not terrible, but not great.

Implemented with Enterprise Customers across region. Very strong Fault Management tool , Extra ordinary correlation & alarming. Not to difficult to implement but eventually adv. correlation is not easy to implement.

It helps our customer in managing the events , correlations & creation of auto tickets with CA Service Desk. Eventually it helps customer in automation.

Its not performance management tool for Network Performance Management you need to have CA UIM or CA eHealth.

12+

We encounter multiples issues , difficult to list all. Some of them far beyond out of the box implementation. Other CA Support helps us to resolve it.

No

Not at all .

Customer complains it lags for performance management which CA has other solution of it.

4-5 No good technical support from CA

CA NSM, Solarwind

its depend on the scope of project. But initial setup is not difficult for this solution.

We are solution implementer , and we had implemented across various ent. customer.

Different customer has different.

Pricing generally is high but you need aquire multiple products in order to accomplish complete infrastructure management. Cant get everything with CA Spectrum alone when it comes to network monitoring. Netflow is not there too.

Yes

With Spectrum, we are implementing a unified monitoring system. We have also purchased CA APM and Agile Central to complement Spectrum and have an ambitious plan to integrate these systems. This will increase the effectiveness of the IT department and will give us the comprehensive view we want to make better decisions that support our business needs.

Spectrum is very useful for us, because the product lets us monitor and control all the various components of our infrastructure that support our business from an IT perspective.

We are also beginning to implement SOI, in order to view our IT infrastructure from a business perspective. As the boundary between business and IT become more confusing, it becomes more valuable to our users that we in IT can easily and quickly understand the impact of IT services within the business. SOI and Spectrum allow us to achieve this. We can optimize the IT services we offer our users and get a quick view of the business impact. Then, we can adjust priorities and solve problems quicker.

From the users perspective the looks and feel of dashboards of Spectrum would be easier to use and for understanding as APM will be in 10.5 version which we saw in CA World.

Also, from our point of view, It´s necessary that CA people in charge of Spectrum tool, work more in order to better integration between Spectrum, with APM, SOI and UIM.

We had some stability issues at the first, but CA resources in Argentina assisted us in to solving these problems. They also helped to fine tune our configuration to support the solution. In the early stages, we implemented a simple configuration of Spectrum, so sometimes we faced up-time problems. But that was due, in part, to the design we implemented. So, now we have implemented a more robust design with more servers and more robust databases. The solution is now stable enough to support the challenges we want to monitor.

We have not experienced any scalability issues, because we used a step-by-step approach. First, we implemented Spectrum at a pilot with branches. We use a dashboard that allows us to monitor the components at our 110 branches to support our products. Next, we applied the same idea to our strategic and critical challenges in website and mobile applications, as well as in our retail and corporate business.

Customer Service:

Very Good, from CA and from Partners

Technical Support:

Technical support is very good. In Latin America, they have good technicians, particularly in Argentina, Chile and Brazil. Whenever we need support on a difficult to solve issue, CA brings in specialists to solve the problem quickly.

Before choosing Spectrum, we used an open-source solution called Nagios. We also used a monitoring tool from BMC called PEM, Patrol Enterprise Management. In 2009, our bank decided to implement the complete suite of tools for IT management and delivery from CA. It's a strategic decision that we took six or seven years ago and we're happy with the switch.

Initial setup was easy, but in the initial stages we need more support from CA and our CA partner in order to understand the capabilities of Spectrum and how to take advantage of all of this.

Yes, from IBM

When we evaluate vendors we look at technical skill and background. We want them to be able to understand our business objectives and our needs. This is the most important. We want them to be more service oriented than product oriented, because the main objective of the IT department in a bank is to maintain up-time regardless of the challenges we face. This is critical to our business.

It's a mature product and has been around for a long time. The technical support is up to speed and it has been well designed.

It gives us a visual representation of what's going on in the network. It hasn't helped me personally but it has helped the network. We haven't moved on to the latest version yet but integration with UIM will be one of it's benefits.

It is not so much in regards to the product itself. In my opinion, it's the documentation that I find is not up to the mark. Instead of reading 400 pages, if they could just give me one simple example, it would make things a lot easier because it's a very complex product that does a lot of complex things. I would like to have illustrations on how to do the simple things as opposed to sending me a link that refers to 400 pages of a manual; it doesn't help. Saying “select the criteria” is meaningless if I don’t have the basic knowledge of what the criteria means or why I’m selecting it.

It is very stable.

The product is very scalable as well.

We have used technical support. It's hit or miss. Sometimes you get people who are really good and sometimes you don't. However, you can usually escalate quite quickly.

Our company was partially using CiscoWorks before this product.

I would definitely recommend this product to others.

The most important criteria whilst selecting a vendor is reliability.

The ability to receive and correlate alarms and determine impacts when outages occur.

As a user of it, it's been fantastic just because it does so many things all in one configuration section. The alarms monitoring of the correlation of events is valuable. It’s been a great experience.

I would like to see easier use in some things, such as trying to configure certain correlations I've been working on. It's been a little bit of a hassle, because I'm not super familiar with their products and the best way to go about things.

Stability has been great once you get it up and running. I've had issues getting new installs to work, but once it's finally going and all set up to go, it's been fantastic.

I was involved in a couple of installations. It was very straightforward. It was just a matter of installing the package, doing the extraction, running it, and it seemed to work.

In terms of choosing a network management applicationsvendor, look for stability and ease of use. I've seen that with the other applications. We have issues happening all the time, although I haven't seen it with CA. I would suggest getting your hands on a product in a lab and play around with it.

• NCM
• Discovery: Scheduled, on-demand, synchronized with eHealth
• REST API
• Spectrum integration with CABI, eHealth, SOI, and EEM
• Modelling of custom device models
• MRTG graphs in real time for interfaces
• Dynamic Global Collections
• Distributed landscapes
• OCC console client
• Web UI console
• Integrated GIS maps
• Self-certification
• Condition correlation
• Event configuration
• OOB Probable Cause database
• MPLS Transport and VPN managers
• Watches and thresholds

CA Spectrum allowed our client, a Tier 1 TELCO, to manage their E2E network infrastructure. The distributed architecture and eHealth integrations enabled a powerful solution for providing Hosted Managed Services to an Enterprise customer. It proactively monitored Enterprise: GMPLS, IPVPN, ISDN, LTE and Leased Line services against set SLAs and maintained a top quality of service.

The easy integration of topology views to a web portal allowed the client to expose actual real time performance and health status of enterprise infra setups. This greatly benefited the client by providing a "wow" factor to their consumers/customers.

• NCM: Needs to be more user-friendly and refined. The ability to quickly see changes across various time configured backups helps to quickly identify changes which could be a root cause to a prevalent problem due to incorrect configurations.
• MRTG graphs need to be more intuitive.
• The GIS map feature needs to be enhanced and synced with topology views of containers and global collections.
• The probable cause database needs to be enhanced.
• As Cisco and Alcatel lose out to cheaper competition like Huawei, ZTE, etc., device certifications need to be updated.

We have been using this solution for four years.

Scheduled discoveries with modelling of large numbers of multi-element devices like switches, MUX, SAN, IP-DSLAM used to freeze up the SpectroServer in r10, although such issues were not faced in r10.1.

We didn't encounter any issues with scalability.

CA Spectrum Technical support is very good, especially the ready availability of extremely skilled technicians across the globe, 24/7.

I am not aware of any previous solutions.

The setup of the distributed SpectroServer is complex for large environments, including the integrations with distributed eHealth environments.

It is very pricey. With the current trend of enterprises moving towards cheaper and more modern looking solutions like ManageEngine, SevOne, ServicePilot, and Paesller.

CA needs to cut down its pricing strategy.

HPE OV and CA IM2.0 were tested, but were found either lacking in scalability or user-friendliness, including the lack of a modern, crisp UI.

It is a great product. It is easy to setup if technically strong resources are available. It reduces the cost of underlying hardware due to the higher monitoring capability per server. Always stick to a distributed approach as it helps in fault tolerance due to the unavailability of HA/DR.

1. NCM – Network Configuration Manager – It is very helpful in maintaining the configuration backups of network devices in a multi-vendor ecosystem involving large number of devices. Device configuration changes can be identified on real time and helps administrators to identify unapproved changes and recover from faulty changes leading to device downtime quickly. Reference configurations are helpful in setting a device configuration and naming convention benchmarks and enables admin teams to work without dependency on resources, as the proper and standardized configurations can always be referred to during making changes to devices.

2. Discovery – CA Spectrum’s discovery utility allows IT/NOC teams to find and model the network and view the topology in the Universe and Global Collections views. Bulk discovery is possible to be run using seed files providing the device IP address and using pre-configured policies containing the list of SNMP community strings. This way large networks can be discovered and modelled in one go. It allows us the flexibility to choose the elements to be modelled and how to model them. The inventory data can be exported in desired file format for further review and distribution to key stakeholders. Device changes can be re-discovered into CA Spectrum using the scheduled discovery feature, ensuring that device changes do not place a device model in risk of getting out of live monitoring and avoiding incorrect alarms.

3. Spectrum Integrations –

a. CA BI – CA Spectrum uses CA Business Intelligence (CABI) for reporting purposes. The flexibility offered by CABI to design user defined reports and display data in a variety of info graphical formats lets admins design reports which can display to management in a glance the state of entire network/ sections of the infrastructure. CABI leverages the powerful SAP BO technology which offers a complete suite of information management, reporting, query and analysis tools in a user friendly and intuitive UI.

b. eHealth – CA Spectrum – eHealth integration offers functionality to map existing eHealth elements with the Spectrum device models to allow eHealth generated performance based alarms to be shown on Spectrum alarm dashboard against the relevant device element model. This creates a single pane of glass view for the NOC teams to view all fault and performance alarms in a single dashboard and helps in reporting and carrying out triages to determine RCA of faults. Also, detailed OOB designed eHealth reports are made available to users of Spectrum on a single click and discoveries can be scheduled on eHealth using only CA Spectrum system. Thus, it eases the system administrator’s job while providing a in-depth analysis of the entire network’s state and performance in real time.

c. SOI – Integrating CA Spectrum with CA SOI is now easy using the CA SOI connector. It helps to collect CA Spectrum data into CA SOI, convert it to a common format and expose the data for visualization, analysis,, management and reporting to organizations senior management. It helps to get a management view on a distributed and large network into a single dashboard view including Cis and alerts consolidated into a single frame. The entire infra can be broken down into logical business services and thus impact of faults can be analyzed on services running / hosted / provided by the underlying infrastructure easily.

d. EEM – This integration enables SSO feature for CA Spectrum. Single Sign-On (SSO) lets CA Spectrum users integrate with LDAP (Lightweight Directory Access Protocol) and Active Directory user management systems.

e. API – CA Spectrum provides web based and REST API for integrations. We found it very useful in creating a custom solution to monitor dynamic IP based devices. It can also be used in embedding CA Spectrum information in web sites or third party apps and fetching data into CA Spectrum especially updating custom attributes of device models with data from external systems.

4. Dynamic Global Collections – Global Collections allow us to model network views from any modelled elements in Spectrum. By specifying rules based on device naming convention/IP address/device type etc. we can dynamically add models into various global collection views as and when they are discovered and modelled in Spectrum. These views are very helpful in breaking down the entire network into logical segregated views based on any criteria which fulfills your business use case.

5. Interface Information - You can view information about the interfaces of a device model by selecting the model and selecting the Interfaces tab in the Component Detail panel. The Interfaces tab displays a list of the configured interfaces and sub interfaces for the selected device, along with the parameters of the interface models. The % utilization (In/Out), packet, errors/discards (%) rates can be viewed in graphical format and thresholds can be set to generate alarms.

The product is rich in features, so a lot of them are valuable.

Network Configuration Manager - Backup network device configuration, enforce policies, enforce configuration and correlate configuration changes with network status. Also has an approval workflow for changes on devices.

Certification - You can certify devices in several levels. The most complex ones give you great results for your efforts, such as a certification using a Model Type Editor for a customized model type along with Event Admin to receive alerts from another source and redirect to your new models based on Name (this allows for creating virtual environments to monitor devices without IP if you have a manager with IP/SNMP for them) or IP, creating Dynamic Alarm Titles. This gives great value, because it becomes really easy for your operations team to know exactly what is happening by just looking at the alarm console, instead of searching for the problem in the alarm description.

Event Correlation - Great to map events and to correlate them before generating alarms. It works with event pairs, event count, event series and other methods; really useful.

Condition Correlation - Great for correlating alarm conditions with devices. For example, you have a router redundancy, and you don’t want two different alarms for Device Down. You can create a domain and rules to specify that if only both of the routers are down, it will generate a new alarm with the possibility to use whatever name you want for that new alarm.

Service Dashboard - This one is particularly great, because customers love it. It allows you to give them a service-oriented view of their infrastructure without having to use another tool for that. Because it’s part of CA Spectrum, the service is affected immediately after an alarm occurrence.

It allowed us to move from an open-source solution using six servers to an enterprise solution using two servers to monitor the same number of elements. Also, of course, the product offered a lot of features ready to use that we didn’t have before.

The Systems Performance Agent is powerful, but, for example, if you want to run a sql query and set a threshold against the returned value, you can’t directly. You need a script with the threshold set in the script and an exit different than zero (0) to generate an alarm. This is something that almost every monitoring solution on the market has built-in. CA UIM, for instance, does that really well. What you have ready to use is the ability to execute the query and get the value for how much time it took to execute.

I have used it since January 2009.

I have not encountered any stability issues so far.

I have not encountered any scalability issues so far.

Technical support is very good.

We previously used a different solution, and we switched mainly because the company needed scalability.

We were using Zabbix, a similar product, at the time, and we had a custom poller written in C and six servers to make it run. However, to get it past 12k elements (the break point for this solution), we would have to redo the entire file system using GFS from Red Hat to allow simultaneous write for a shared storage area. So, when we presented the project, the company realized it was time to invest in an enterprise solution. CA Spectrum was the winner after a series of tests and PoC process with the suppliers.

To install CA Spectrum and discover the network, it required half a day, back in 2008. Some of the competitors couldn’t complete the PoC after one week.

And CA Spectrum offered also a variety of OOTB features such as an easy certification process, VLAN discovery and other stuff.

Initial setup was very simple; as simple as Next - Next – Finish. The complex part isn’t the setup, but the planning (architecture and such).

Right now, it’s device based, so count what you have and estimate your total growth for at least one year. CA works on a trust-based method.

At the time we decided to switch from an open source to an enterprise solution, we evaluated HP OpenView, IBM Tivoli, EMC Smarts.

Don’t try to do the planning without consulting a CA Partner or an expert. You will need help in other matters also; must have a good knowledge of SNMP protocol. Always read the documentation, which is very rich, and use the communities for extra help besides the support.

CA Spectrum is awesome for networks, but when paired with CA UIM, you can have everything: datacenter, networks, virtualization and on.