SAN requirements and imaging facilities (acceptable server downtime)

[Alessandro Esposito]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear all,
    in order to guarantee backups, desktop portability and integration f various
advanced IT services, our institute established a SAN and virtual environments
where all IT services are centralized.

This works wonderfully but at the cost of downtimes caused by planned
(which is not an issue) and unplanned downtime of all services.

The obvious solution is to make research instruments independent from the
main server with loss of some services and increase of maintenence cost on
individual computers.

Have anybody done a serious analysis of SAN requirements in order to maintain
continuity of service? We are an institute serving between 150-200 employees
and we are discussing on possible improvment to our infrastructure.

I guess the most important question I am trying to answer is: 1) how much
unplanned downtime is considered acceptable by other research institutes (in
particular those with significant imaging facilities) and 2) if some key solutions
have been identified for minimizing the impact on running microscopes during
an IT outage.

Again, although the basic solution is to isolate instruments, I hope we could
discuss about a better compromise :)

Cheers,

Alessandro

[Chris Tully-2]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Alessandro,

This response in based not on personal experience but on helping a major
hospital network setup a regional digital pathology network covering 5
hospitals separated by anywhere from 5 to 200 miles, while I was working
for Aperio.

For many reasons starting with the volume of data they expected to
generate (multiple GB per day), they built not one but two SANs at two
different sites.  Then every connected instrument was configured to
consider one of the sites (A) as it's primary, with automatic fail over
to the secondary site (B). SAN site B was configured to live mirror site
A so that in the event that A went down the most anyone would loose is
the slide that was actively being scanned when the fail over happened.
Further the scanning software was designed to recognize this failure and
flag it for future correction.

Obviously, mirrored SAN systems become expensive, but if minimal down
time is required (desired) then that is the only solution I can come up
with. Check into each instrument connected to your SAN. As with the
Aperio Scanners that I was helping to implement, your instruments may
have a third fail over mode either automatically setup of available to
be configured for storing data locally when the SAN is unavailable. The
challenge then becomes migrating data from local storage to the SAN once
it is available again.  You will want to push the manufacturers of each
instrument to work with your IT department to automate that migration as
much as possible (ideally, the SAN or the instrument will detect the
restored connection and automatically push/pull the locally stored
data). Then your only remaining challenge is to talk with your IT people
and find out how long it would take them to recover from a worst case
scenario and add enough local storage to each instrument to cover that
event.

How far you go down that path will obviously be governed by budget,
backed up by need.  The more failure modes you can plan for the less
down time your users will experience.

Obviously having data stored locally for some time will not be as
convenient, but at least it is not lost.

Chris Tully

On 9/18/2013 3:37 AM, Alessandro Esposito wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Dear all,
>      in order to guarantee backups, desktop portability and integration f various
> advanced IT services, our institute established a SAN and virtual environments
> where all IT services are centralized.
>
> This works wonderfully but at the cost of downtimes caused by planned
> (which is not an issue) and unplanned downtime of all services.
>
> The obvious solution is to make research instruments independent from the
> main server with loss of some services and increase of maintenence cost on
> individual computers.
>
> Have anybody done a serious analysis of SAN requirements in order to maintain
> continuity of service? We are an institute serving between 150-200 employees
> and we are discussing on possible improvment to our infrastructure.
>
> I guess the most important question I am trying to answer is: 1) how much
> unplanned downtime is considered acceptable by other research institutes (in
> particular those with significant imaging facilities) and 2) if some key solutions
> have been identified for minimizing the impact on running microscopes during
> an IT outage.
>
> Again, although the basic solution is to isolate instruments, I hope we could
> discuss about a better compromise :)
>
> Cheers,
>
> Alessandro

--
*Chris Tully*
Principal Consultant
240-475-9753


      Image Incyte, LLC

<http:%5C%5Cwww.ImageIncyte.com> [hidden email]
<mailto:[hidden email]>

[Mark Cannell-2]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

There are some simple tools to achieve folder mirroring e.g. http://www.techsoftpl.com/backup/index.php. For Unix systems this could be done by rsync or scripts called by CRON.

HTH

Cheers

On 18/09/2013, at 2:08 PM, Chris Tully <[hidden email]> wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Alessandro,
>
> This response in based not on personal experience but on helping a major hospital network setup a regional digital pathology network covering 5 hospitals separated by anywhere from 5 to 200 miles, while I was working for Aperio.
>
> For many reasons starting with the volume of data they expected to generate (multiple GB per day), they built not one but two SANs at two different sites.  Then every connected instrument was configured to consider one of the sites (A) as it's primary, with automatic fail over to the secondary site (B). SAN site B was configured to live mirror site A so that in the event that A went down the most anyone would loose is the slide that was actively being scanned when the fail over happened. Further the scanning software was designed to recognize this failure and flag it for future correction.
>
> Obviously, mirrored SAN systems become expensive, but if minimal down time is required (desired) then that is the only solution I can come up with. Check into each instrument connected to your SAN. As with the Aperio Scanners that I was helping to implement, your instruments may have a third fail over mode either automatically setup of available to be configured for storing data locally when the SAN is unavailable. The challenge then becomes migrating data from local storage to the SAN once it is available again.  You will want to push the manufacturers of each instrument to work with your IT department to automate that migration as much as possible (ideally, the SAN or the instrument will detect the restored connection and automatically push/pull the locally stored data). Then your only remaining challenge is to talk with your IT people and find out how long it would take them to recover from a worst case scenario and add enough local storage to each instrument to cover that event.
>
> How far you go down that path will obviously be governed by budget, backed up by need.  The more failure modes you can plan for the less down time your users will experience.
>
> Obviously having data stored locally for some time will not be as convenient, but at least it is not lost.
>
> Chris Tully
>
> On 9/18/2013 3:37 AM, Alessandro Esposito wrote:
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> Dear all,
>>     in order to guarantee backups, desktop portability and integration f various
>> advanced IT services, our institute established a SAN and virtual environments
>> where all IT services are centralized.
>>
>> This works wonderfully but at the cost of downtimes caused by planned
>> (which is not an issue) and unplanned downtime of all services.
>>
>> The obvious solution is to make research instruments independent from the
>> main server with loss of some services and increase of maintenence cost on
>> individual computers.
>>
>> Have anybody done a serious analysis of SAN requirements in order to maintain
>> continuity of service? We are an institute serving between 150-200 employees
>> and we are discussing on possible improvment to our infrastructure.
>>
>> I guess the most important question I am trying to answer is: 1) how much
>> unplanned downtime is considered acceptable by other research institutes (in
>> particular those with significant imaging facilities) and 2) if some key solutions
>> have been identified for minimizing the impact on running microscopes during
>> an IT outage.
>>
>> Again, although the basic solution is to isolate instruments, I hope we could
>> discuss about a better compromise :)
>>
>> Cheers,
>>
>> Alessandro
>
>
> --
> *Chris Tully*
> Principal Consultant
> 240-475-9753
>
>
>     Image Incyte, LLC
>
> <http:%5C%5Cwww.ImageIncyte.com> [hidden email] <mailto:[hidden email]>

Mark  B. Cannell Ph.D. FRSNZ
Professor of Cardiac Cell Biology
School of Physiology &  Pharmacology
Medical Sciences Building
University of Bristol
Bristol
BS8 1TD UK

[hidden email]

[Alessandro Esposito]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Chris,
    thank you for your interesting example. I did not consider at all the
possibility of the redirect to be handled by the microscope in itself and I
should check if this is supported at all by a typical microscope (probably
not). It would be smart that if in the absence of the SAN, the
micropscope would switch saving to the local harddrive. I'll look into it.
For other systems, such as screening platforms, that would be not
possible and the solution you mention should be adopted, but
manufacturers could implement saving locally with metadata the would
be later recovered and transfered to the SQL server togehter with the
data on the SAN.

Thank you

Alessandro


On Wed, 18 Sep 2013 09:08:12 -0400, Chris Tully
<[hidden email]> wrote:

>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>Alessandro,
>
>This response in based not on personal experience but on helping a
major
>hospital network setup a regional digital pathology network covering 5
>hospitals separated by anywhere from 5 to 200 miles, while I was
working
>for Aperio.
>
>For many reasons starting with the volume of data they expected to
>generate (multiple GB per day), they built not one but two SANs at two
>different sites.  Then every connected instrument was configured to
>consider one of the sites (A) as it's primary, with automatic fail over
>to the secondary site (B). SAN site B was configured to live mirror site
>A so that in the event that A went down the most anyone would loose
is
>the slide that was actively being scanned when the fail over happened.
>Further the scanning software was designed to recognize this failure
and
>flag it for future correction.
>
>Obviously, mirrored SAN systems become expensive, but if minimal
down
>time is required (desired) then that is the only solution I can come up
>with. Check into each instrument connected to your SAN. As with the
>Aperio Scanners that I was helping to implement, your instruments may
>have a third fail over mode either automatically setup of available to
>be configured for storing data locally when the SAN is unavailable. The
>challenge then becomes migrating data from local storage to the SAN
once
>it is available again.  You will want to push the manufacturers of each
>instrument to work with your IT department to automate that
migration as

>much as possible (ideally, the SAN or the instrument will detect the
>restored connection and automatically push/pull the locally stored
>data). Then your only remaining challenge is to talk with your IT people
>and find out how long it would take them to recover from a worst case
>scenario and add enough local storage to each instrument to cover that
>event.
>
>How far you go down that path will obviously be governed by budget,
>backed up by need.  The more failure modes you can plan for the less
>down time your users will experience.
>
>Obviously having data stored locally for some time will not be as
>convenient, but at least it is not lost.
>
>Chris Tully
>
>On 9/18/2013 3:37 AM, Alessandro Esposito wrote:
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> *****
>>
>> Dear all,
>>      in order to guarantee backups, desktop portability and

integration f various
>> advanced IT services, our institute established a SAN and virtual
environments
>> where all IT services are centralized.
>>
>> This works wonderfully but at the cost of downtimes caused by
planned
>> (which is not an issue) and unplanned downtime of all services.
>>
>> The obvious solution is to make research instruments independent
from the
>> main server with loss of some services and increase of maintenence
cost on
>> individual computers.
>>
>> Have anybody done a serious analysis of SAN requirements in order
to maintain
>> continuity of service? We are an institute serving between 150-200
employees
>> and we are discussing on possible improvment to our infrastructure.
>>
>> I guess the most important question I am trying to answer is: 1)
how much
>> unplanned downtime is considered acceptable by other research
institutes (in
>> particular those with significant imaging facilities) and 2) if some
key solutions
>> have been identified for minimizing the impact on running
microscopes during
>> an IT outage.
>>
>> Again, although the basic solution is to isolate instruments, I hope
we could

>> discuss about a better compromise :)
>>
>> Cheers,
>>
>> Alessandro
>
>
>--
>*Chris Tully*
>Principal Consultant
>240-475-9753
>
>
>      Image Incyte, LLC
>
><http:%5C%5Cwww.ImageIncyte.com> [hidden email]
><mailto:[hidden email]>

[mcammer]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

We've thought about this and decided that regardless where data ultimately reside, all research instruments are stand alone when collecting data.  Regardless of network status, they may collect data uninterrupted.  

We can tolerate a wait to recall data residing on a server that is temporarily down but we cannot tolerate interruption of ongoing experiments, especially those involving live material.


Regards,

Michael


________________________________________________________
Michael Cammer, Assistant Research Scientist
Microscopy Core, NYU Langone Medical Center & Skirball Institute of Biomolecular Medicine
Cell: [phone]   Microscopy Lab: [phone]   Dustin Lab: [phone]  

[Cameron, Lisa]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hello -

Have people tested and found optimal 2-photon laser excitation wavelengths for some of the newer fluorophores like mTurquoise2 and Clover?

I'd be interested to hear what people have tried / found.

Our specific application is FLIM-FRET.
We do the measurements with a Becker & Hickl time domain system on a Zeiss 710 with 2p laser.

Thank you for any information -
Lisa

---------------------------------------
Lisa Cameron, Ph.D.
Director of Confocal and Light Microscopy Core
Dana Farber Cancer Institute
450 Brookline Ave.; JF 621
Boston, MA 02215
http://microscopy.dfci.harvard.edu


The information in this e-mail is intended only for the person to whom it is
addressed. If you believe this e-mail was sent to you in error and the e-mail
contains patient information, please contact the Partners Compliance HelpLine at
http://www.partners.org/complianceline . If the e-mail was sent to you in error
but does not contain patient information, please contact the sender and properly
dispose of the e-mail.

[Tobias Rose]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Indeed - I'd like to second that question.

Right now I'm using 850nm for mTurq2. This may be suboptimal, though.

Well: I should probably just test it.

Tobias


-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Cameron, Lisa
Sent: Friday, September 20, 2013 15:40
To: [hidden email]
Subject: 2-photon excitation wavelength recommendations for mTurq2 and clover

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hello -

Have people tested and found optimal 2-photon laser excitation wavelengths for some of the newer fluorophores like mTurquoise2 and Clover?

I'd be interested to hear what people have tried / found.

Our specific application is FLIM-FRET.
We do the measurements with a Becker & Hickl time domain system on a Zeiss 710 with 2p laser.

Thank you for any information -
Lisa

---------------------------------------
Lisa Cameron, Ph.D.
Director of Confocal and Light Microscopy Core Dana Farber Cancer Institute
450 Brookline Ave.; JF 621
Boston, MA 02215
http://microscopy.dfci.harvard.edu


The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.