[Info-vax] A Samba alternative, could this be something for VMS?

[David Froble]

Kerry Main wrote:
>> -----Original Message-----
>> From: Info-vax [mailto:info-vax-bounces at rbnsn.com] On
>> Behalf Of IanD via Info-vax
>> Sent: 06-Sep-16 9:08 PM
>> To: info-vax at rbnsn.com
>> Cc: IanD <iloveopenvms at gmail.com>
>> Subject: Re: [Info-vax] A Samba alternative, could this
> be
>> something for VMS?
>>
>> On Wednesday, August 31, 2016 at 11:15:04 PM UTC+10,
>> Kerry Main wrote:
>>
>>> Bottom line - any new technology needs to adopt to
>> meet business DR/DT/RTO/RPO SLA's - not the other way
>> around.
>>>
>>> Regards,
>>>
>>> Kerry Main
>>> Kerry dot main at starkgaming dot com
>> +1
>>
>> Business availability is #1 priority where I am,
> everything is
>> skewed around that ideal
>>
>> VMware has done well in terms of the business spend
>> here because they perceive VMware as giving them total
>> availability
>>
>> Our OpenVMS systems were clustered across different
>> geographical regions and that came with issues of it's
> own.
>> Stretched vlan's are the focal hate point of networks.
>> Latency on disk writes was somewhat annoying but the
>> old SCSI HSZ disks were by far the bigger pain than
> going
>> across a vlan. Being a Telco, they threw plenty of money
>> into hefty links (this was spanning three DC's after
> all).
>> When the business decided geographical redundancy
>> wasn't worth it anymore because the new DC was
>> supposedly designed with a near 100% up-time
>> guarantee, our world got a lot smaller as we
> consolidated
>> to a single DC
>>
>> Charon with flash drives replaced the HSZ based 8400 and
>> I/O rates literally went through the roof. SAN is now
> the
>> bottleneck (one system still remains on physical
>> hardware)
>>
>> Redundancy in the geographical form for our business has
>> been replaced by pushing that responsibility onto the DC
>> to manage. Supposedly the DC is designed to take a plane
>> strike and other such mad things human being want to do,
>> so the concept of clusters spanning long distances for
> us
>> went away (whether right or wrong is not for me to say)
>>
>> I think when we see those blazingly fast memory based
>> systems, the business will embrace them endlessly and in
>> the stampede for performance, all concerns about
>> housing everything on a single bit of hardware will be
>> thrown to the wind until something happens and they get
>> bit firmly in the arse and then suddenly redundancy will
>> come into sharp focus again
>>
>> Human being have a psychological flaw in that they are
>> overly optimistic (Can't remember the actual research
>> article now)
>>
>> <sarcasm>
>> Oh look, there's some new technology over there. Looks
>> good, but what about...Stand out of the way or you'll
> get
>> trampled by the mob...
>> </sarcasm>
>>
> 
> Re: consolidate to one DC ... really bad idea.
> 
> Even if it is a tier 4, all it takes is for one person
> with SARS or some other exotic pandemic issue to walk into
> the front desk area of the DC and that's it - when the
> guys in white suits find out, everyone in that entire
> facility physically goes home for 10 days.
> 
> This is where you find out if you can reboot, power off/on
> and if you have enough remote access / VPN lines.
> 
> If a critical system crashes - too bad, the Maint folks
> won't be able to get in until the quarantine is over.
> 
> Question - if no one can get into the lone DC for 10 days,
> what would be the impact on your company?
> 
> And I have pointed this out before, but this actually
> happened to HP in Toronto during the SARS epidemic: (mind
> you they had dual site strategy, so the impact was not as
> bad)
Yep!  Doesn't even require a small asteroid ....