Understand the trade-offs with reactive and proactive cloudops

ByJosephine J. Romero

May 23, 2022 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,
Understand the trade-offs with reactive and proactive cloudops


It is a no-brainer. Proactive ops devices can determine out problems in advance of they turn out to be disruptive and can make corrections without the need of human intervention.

For occasion, an ops observability software, such as an AIops tool, sees that a storage process is generating intermittent I/O mistakes, which implies that the storage method is possible to experience a key failure sometime shortly. Data is instantly transferred to one more storage system using predefined self-healing processes, and the program is shut down and marked for maintenance. No downtime happens.

These sorts of proactive processes and automations occur countless numbers of situations an hour, and the only way you’ll know that they are operating is a lack of outages caused by failures in cloud solutions, apps, networks, or databases. We know all. We see all. We keep track of information in excess of time. We take care of issues prior to they develop into outages that damage the business.

It is good to have this technology to get our downtime to close to zero. Nevertheless, like anything, there are excellent and undesirable factors that you will need to take into account.

Traditional reactive ops engineering is just that: It reacts to failure and sets off a chain of activities, together with messaging human beings, to proper the challenges. In a failure event, when a little something stops functioning, we quickly understand the root cause and we take care of it, either with an automatic method or by dispatching a human.

The downside of reactive ops is the downtime. We typically don’t know there is an challenge right until we have a complete failure—that’s just portion of the reactive procedure. Generally, we are not monitoring the facts all over the source or provider, these types of as I/O for storage. We emphasis on just the binary: Is it performing or not?

I’m not a lover of cloud-based mostly method downtime, so reactive ops would seem like something to stay clear of in favor of proactive ops. Nonetheless, in several of the scenarios that I see, even if you have purchased a proactive ops resource, the observability devices of that resource may perhaps not be able to see the information necessary for proactive automation.

Big hyperscaler cloud companies (storage, compute, databases, synthetic intelligence, and so forth.) can observe these techniques in a great-grained way, this sort of as I/O utilization ongoing, CPU saturation ongoing, etc. A great deal of the other technological innovation that you use on cloud-dependent platforms may only have primitive APIs into their inner operations and can only convey to you when they are doing the job and when they are not. As you might have guessed, proactive ops equipment, no make a difference how fantastic, will not do significantly for these cloud resources and providers.

I’m obtaining that much more of these sorts of programs operate on public clouds than you could possibly feel. We’re paying out big bucks on proactive ops with no capacity to watch the interior systems that will provide us with indications that the methods are most likely to fail.

Moreover, a general public cloud source, these as big storage or compute units, is currently monitored and operated by the provider. You are not in command more than the means that are offered to you in a multitenant architecture, and the cloud suppliers do a pretty great job of furnishing proactive functions on your behalf. They see troubles with components and software program methods very long in advance of you will and are in a considerably superior position to resolve factors prior to you even know there is a challenge. Even with a shared obligation model for cloud-centered means, the providers acquire it upon on their own to make sure that the expert services are working ongoing.

Proactive ops are the way to go—don’t get me improper. The trouble is that in lots of occasions, enterprises are generating big investments in proactive cloudops with very little skill to leverage it. Just saying.

Copyright © 2022 IDG Communications, Inc.


Resource hyperlink