fail over

[Chris Waigl]

"Failover" is a familiar term in IT system management jargon, and I've seen it (increasingly?) used metaphorically in more general contexts.

Failover (or fail-over) can be a noun or a verb. It means something quite different from "backup" in that it refers to the situation in which a service, when being affected by an outage, is automatically switched over to a stand-by or redundant system and therefore continues to be offered with little or no visibility to the users of the service that an outage has occurred. A stand-by system would be one that is substantially inactive when no outage occurs whereas a redundant system could be one that works in a shared or load-balanced configuration with the one that encounters the outage. [It occurs to me that a non-technical speaker may consider this situation an instance of "back-up", but in the technical sense, (data) back-up refers to redundant copies of data or system snapshots being taken on a pre-defined schedule and stored in a location different from the original data or system as well as provisions to restore the original system's state from those copies.]

As a verb, failover can be used transitively and intransitively: "The power supply unit overheated and failed over to the secondary power supply unit." "During the weekly system testing, all database servers were  failed over [to some stand-by database server or other]."

You would find the word failover surrounded by related jargon terms such as availability, uptime, redundancy, robustness, SLA (service-level agreement: a contractual obligation of a service provider that sets out minimally acceptable values for uptime and/or severity of outages) etc. etc.

Chris Waigl

On 11 May 2011, at 03:22, Victor Steinbok wrote:

> I thought I had sent a message on this last week, but, apparently, I did
> not include correct address.
>
> Below, discussion of "fail over" as a new term, question about
> figurative usage of "radio silence" and antedating of "mission-critical"
> (1976-->1968-->?) [emphasis added in quotations throughout]
>
> From a message on ABA eDiscovery committee:
>
>> There appears to be an interesting eDiscovery/preservation issue that
>> stands apart from the business-needs modeling (high availability,
>> *failover*, service level guarantees) discussed by this linked
>> article.
>>
>> What do I mean?  That term *"failover"* may likely be synonymous
>> with...backup. So, where there is discoverable, relevant, ESI in the
>> cloud, there is also backup or, some degree and form of replicated
>> data sets seeding the same, or other clouds.
>
> I initially interpreted that as a noun. I appear to have been
> mistaken--it's both noun and verb. Here's the text from the original CRN
> slide:
>
> http://fwd4.me/00lM
> Slide 3
>> "It will force customers to ask more questions, put a disaster
>> recovery plan in place that includes more than one provider for
>> example, or *fail over* into another region. Service providers are in
>> a tough spot now to do what they should have been doing since the
>> beginning, and that's educating their customers with all the options
>> available to them."
> Slide 4
>> "The exposure here is that when leveraging the cloud, the buyer needs
>> to fully understand the technology and the SLAs that each cloud
>> provider offers. High availability and data center *failover* are
>> offered at different levels. Clients need to fully understand what
>> they are signing up for, but also what their tolerance is for each
>> system or environment that is being migrated to the cloud."
>
> There is an assortment of cloud-related terminology as well, but these
> appear to be fairly typical:
>
>       cloud service, cloud charlatans, pseudo-cloud, etc.
>
> Also, from Slide 11:
>
>> Amazon came under criticism for its *lack of communication* during and
>> following the cloud outage, and *radio silence* won't fly when customer
>> infrastructures aren't running as promised.
>
> "Radio silence" is interesting here--not because it's new, but because
> the OED is vague on what constitutes "extended use":
>
>> radio silence n. abstention from radio transmission; absence of radio
>> transmission; also in extended use.
>
> The four examples don't help--the first three deal with the traditional
> jargon and the fourth is, at best, ambiguous. Here we have something
> that is a completely unambiguous transferred usage--radio silence==lack
> of communication. Is there a need for elaboration?
>
> One of the slides mentions "mission critical"--something that OED dates,
> fairly generically, to 1976. The date is off target and the origin of
> "mission critical" appears to have been with NASA.
>
> In particular, I managed to date one snippet somewhat definitively:
>
> http://goo.gl/OlX2G
> Appolo Accident: Hearings Before the Committee on Aeronautical and Space
> Scienves. United States Senate. 1968
>> Action.--Mission-critical positions have been identified and
>> incumbents are being submitted for required investigation.
>
>
> It's rather obvious that this is not the earliest citation, but the GB
> material from the 1960s, especially government reports and papers from
> RAND and JPL are so badly mistagged that untangling it all is nearly
> impossible. All the RAND papers appear to have either completely wrong
> dates or dates that correspond only to the earliest in the collection.
> There are some hits that show 1966, 1964, 1962, but they are unverifiable.
>
> Certainly some targeted searches within documents dealing
> with the space program some time between 1960-1968 will give more
> snippets and, perhaps, some history of the term. It may go back even
> further. But that's a job for another time--and for someone else.
>
>       VS-)
>
> ------------------------------------------------------------
> The American Dialect Society - http://www.americandialect.org

--
Chris Waigl -- http://chryss.eu -- http://eggcorns.lascribe.net
twitter: chrys -- friendfeed: chryss

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org