Tools Upgrade - downtime required.....?

[johndanter]

Hi List,

We are on E9.0 TR 9.1.3.1 and we wish to upgrade to 9.1.5.2.

How much downtime is required? We've been advised 4 days and this is just for DV.

Am I right to assume it's 4 days per environment or is this 4 days per E1 installation.

We can't afford to shut down PROD for 4 days (when we add the TR update there)

thanks

John

 

[clmates]

Hi

We upgraded from 9.1.2.1 to 9.1.4.6 last year, and in that case we used the PY setup to setup also multi foundation as we dont had any server for testing the tools, we tested it in PY and later, we activated it on DV, PD

the downtime was about 1 day for PY setup and multi foundation

the downtime to activate it on PD was about 4 hours


But this can ve very different regarding your installation, servers involved, etc, in our case, we are a small company with only one web server, one enterprise server and the database server

Regards


Regards

 

[jdel6654]

Hi List,

We are on E9.0 TR 9.1.3.1 and we wish to upgrade to 9.1.5.2.

How much downtime is required? We've been advised 4 days and this is just for DV.

Am I right to assume it's 4 days per environment or is this 4 days per E1 installation.

We can't afford to shut down PROD for 4 days

thanks

John

If you have a CNC that knows what they are doing, they can set up multi-foundation without disrupting any users. Assuming that there are no new physical servers involved and the supporting apps don't change (db, web server, OS, etc.), multi-foundation is basically just another directory. Multi-foundation setup is about 1-3 hours per pathcode even with the system up.

Downtime of anything more that a few hours is laughable. Curious to know how / why you were told it would take 4 days.

 

[johndanter]

Hi

Yeah we have multifoundation on here currently as we actually have 2 TRs here ATM. DV and PY and different to UT and PD.
We never got around to adding the DV/PY TR up into PROD, so this is just another attempt at a more recent TR

There are no new servers no, all current architecture to be used.

I'm curious too
I'll find out and post here

 

[altquark]

4 days outage is outrageous for a Tools Upgrade. I have no idea why they quoted that number.

A tools upgrade can be set up to be either rolled out "gently" using multifoundation - or you can roll it out to non-production first and then test, and then roll out to production during a scheduled downtime.

Tools Upgrades from a technical standpoint are incredibly easy to roll out - almost as easy as clicking "update" on an iphone ! However, any tools release has to be thoroughly tested with your functionality - especially with interfaces as tools releases might alter how conversion files might look or versions of PDF's, etc etc

So, 4 days of non-production testing might sound more reasonable - but this should absolutely be done without affecting production - followed by a production package build and deployment of the new tools release (maybe a 4-5 hour downtime effort).

Note that the Server Manager portion could be done ahead of time - as well as all server manager agents. I don't feel that Server Manager dramatically affects production/non-production systems very much - and you can host a 9.1.4 tools release, for example, on a 9.1.5.2 server manager. Its also very, very easy to roll back server manager if there are issues.

Every customer should have a plan to keep current on their tools release - preferably at least twice a year for most customers. That way they do not get left behind.