What is most valuable?
Over time we have found that if there are any issues with communication between the various components [ scheduler (eventor) ,app server (as_server) , database (event server), the agents, etc ] that when connectivity is restored, jobs start executing as desired and anything that is stuck or is in an ‘strange state’ is easily found and rectified using the tools provided by autosys.
We have found that most of the time the WorkLoadAutomationAE software requires very little maintenance and has achieved an almost 100% uptime in all the instances we have running in our DEV/TEST/and PROD environments.
IOW – Although there are many components involved, once things are setup correctly, the application performs solidly, is extremely redundant (again, if setup correctly) and has very little maintenance overhead.
How has it helped my organization?
We've just recently upgraded to the 11.3.6 Workload Automation platform. We're still implementing the pieces, but what we found is, there's a lot more reporting with it. It's a lot more stable. There are a lot more features available, that we've yet to even take advantage of.
What needs improvement?
When we first went to 11.3.6, they didn't have their agent monitoring platform produced. As of this date, we haven't even installed their agent orchestrator yet. We're trying to work to get those in. I can't specify what I would like right now, unfortunately, because we haven't fully implemented the product, so I don't know what's missing.
They have a lot of great features; these new job types that we've yet to even start implementing. Some of that's going to require some philosophical change on how we promote the code, so I don’t have hard information on it. I would just say to CA, stay on the roadmap.
My opinion, from what I've seen, is, when we started with 11.3.6 service pack 0, the initial release (now they're up to service pack 5), I've seen them address all my concerns regarding, how do we manage the agents from a centralized location and all? I think they're on the right track, but at this point, I can't say, "Wow, they're missing this key feature.", because they seem to have everything that I've been thinking about in the pipeline.
When we started the new product and we introduced it, we started getting some new alarm types and the agent went offline. The architecture about how the scheduler monitors the agents, it knows when the agent on the workload server actually disappears and it reports that. I was talking to our service engineers and so forth, and I said, "You have an alarm that goes off, that says the scheduler's lost contact with server x's agent and it's logged in a log. When it recovers, that's logged in the log, but there's nothing from an alert perspective." Basically, you have a system that alerts you when there's a problem, but doesn't alert you when it's resolved.
I said, "We don't look at AutoSys, or in this case, Workload Automation, to be the correlation engine for what we escalate, but if you're going to put out an alarm that says the sky is falling, I've lost contact with an agent, and you log that in your log." I kind of said, "It would be nice if you put out the same type of information alert, that shows that it cleared, since the log of the application shows that. You're basically saying, the sky is falling and then when the sky is caught, you don't alert us that it's all better.”
It caused a little confusion since our NOC is reaction based. We see the alert and then when it clears, they don't see that it cleared, so we have to do additional work. I've talked to CA about it. They said, "It's an interesting idea. We'll take it under advisement.", because it's not a bug, it's more like undocumented features.
What do I think about the stability of the solution?
What I like about the product is, we've had issues at the infrastructure level, regarding components that stopped seeing each other, or they got out of sync regarding, say, the database communications. Once the system reestablishes itself, you don't have to worry about if something happens in your infrastructure regarding, “Oh my God, my jobs are going to get out of sync.” The system does a really good job of piecing current status together and giving you accurate results. That's a good stability piece.
What do I think about the scalability of the solution?
Our implementation is not that huge on it. It's a couple of small instances of Workload Automation that do a lot of work. From a scalability perspective, I enjoy the fact that, when we scale our networks – we have our servers that actually perform the work out in disparate parts of the company – we don't have to bring the whole infrastructure in. The application server layer of the product line really allows you to centralize the control, without having to have direct, direct contact to every single component out there. It's a nice tiered approach. I see it that the scalability is in line with where I see our company going.
How is customer service and technical support?
Interestingly enough, with technical support, it's a huge company, so even though you have a one-to-one relationship with, maybe, your professional services or your account manager, when you call support, they don't really have that history. I found that, with a lot of companies, you're concerned about, I have to start with a tier-one person, I have to re-explain my problem.
What I've experienced with CA, over the last, say, 2-3 years that I've actually worked with them directly and their product, is, they understand that. I've actually gotten very good results, because it's not just a technical part that they see. I think that CA and their technical support for this product understands the human condition that's involved, the concerns. I don't have to worry that, if I have an issue, I have to justify my concerns about my support issue. They understand that and they address it proactively, which, for the most part, gives me an overall good customer experience using their support.
How was the initial setup?
We just recently upgraded our Workload Automation. We went from 4.5.1 to 11.3.6. I think we did about an eight-year jump, in one fell swoop. It was nice to be able to help work with their teams to architect this solution in our upgrade, where we actually brought in our business needs. It wasn't just, “Hey, we're just going to replace it and move it forward.” That’s kind of like what we did, because we wanted to keep the solution as transparent to the end users that use it. We worked with the company and came up with something that makes sense for us.
What other advice do I have?
Based on my experience, make sure you ask a lot of questions and advocate for what your business needs are. When you get your solution, investigate it. Don't just say, like CA, like any other vendor says, "This product can do x, y and z." Don't take that at face value. Ask the right questions, because I find that in technology, especially what I experience, when people talk about what my needs are and what my product can do, sometimes they're using the same words with different meanings based on the perception of the user or the context of the conversation.
I'm thinking, be clear, especially if you're on the technical piece. If you're going to be supporting this, ask a lot of questions, because what I found is, after the implementation of this product, it's kind of hard to go back and re-architect it and things like that. Advocate for yourself, but it's a fine product. I advocate for it all the time to everybody, because it does what it needs to do.
I think that the product has so many moving parts and so many features that it's a plethora of great stuff, whether it gets implemented for a specific vendor really depends on the vendor’s needs. I think it's a fabulous product that you can really fine tune for what your specific business needs are. It's scalable, it can grow. You don't have to worry about re-buying a hundred different things.
To me, if you do it right, it's very easy to manage, which is great. My team that I work with, we've been working with this product for eight years now and we were talking about it and somebody says, "The great thing about this product is, when there's an issue, you don't have to worry about picking up all the pieces and figuring out what happened." The system kind of comes back online, makes the reports that you need.
Which version of this solution are you currently using?