Hello Kavin, 
I'm sorry for what you face... I hope if there is any one tell us what is the 
best practice for rolling back.. for me i clone the vm and update it if there's 
problem i delete it and open the master vm 

Sent from Yahoo Mail on Android 
 
  On Sat, Jun 20, 2020 at 8:23 PM, Kevin Creechan<[email protected]> wrote:   Hello,


I am looking for users who may have experience running SOGo in a virtual 
server, specifically for anyone who's had to roll back after an update.

After an update went wrong the other day we had rolled back 20 minutes. 
As a result we started getting calls:

- users receiving emails with phantom recipients added that were never 
on the original email.
- users receiving emails with extra attachments that they'd already 
received earlier from completely different recipients.
- users sending emails to clients with the wrong attachment or an 
attachment that was never attached.
- users sending emails to clients with additional and unrelated 
recipients that were never added.

All users using Outlook Pro Plus 2019 on an EAS connection.

After a lot of hair pulling we noted via Webmail that the emails were 
actually fine, and that the extra items were only happening in Outlook. 
Once we re-added their Outlook profiles and let it re-download their 
mail, everything showed with correct recipients and attachments.

I suspect that Outlook keeps a relational database of email items like 
Subjects, recipients and attachments. Perhaps rolling back the server 20 
minutes resulted in duplicate ID's on things so that newly arriving 
emails would have related items matching one that came in during the 20 
missing 20 minutes. Just theory of course, but seems to fit the evidence.

Can anyone relate, or anyone else provide any advice with regards to how 
to properly roll back a bad update without risking these symptoms in the 
future?



-- 

Thanks,

Kevin


-- 
[email protected]
https://inverse.ca/sogo/lists
  
-- 
[email protected]
https://inverse.ca/sogo/lists

Reply via email to