Upgrade from 9.5U3 to 9.5U4 failed with "Database version downgrade detected"

[Moodie007]

I have now had this issue on multiple installations. Have had to reinstall and reconfigure certain aspects. Was a real mess. I have two support cases open. 03388030 and a slightly older one (24th JAN) 03378235. Still waiting on a full explination/resolution.

[Moodie007]

Okay I have a workaround. I wasn’t through enough when I changed the region originally - I suspect I didn’t reboot after making the changes. You won't need to change all of these settings, no doubt - such as "New User Accounts" tick box, but this is what I did.

I have changed the following Region settings under Control Panel:
Formats Tab - Format: English (United States)
Location Tab - Home Location: United States
Administrative Tab - Change system locale...
English (United States)
Administrative Tab - Copy Settings for Current User, Welcome screen and New User accounts:
Display language: English (United States)
Input language: English (New Zealand) - US
Format: English (United States)
Location: United States
Tick both Tick boxes for "Welcome screen and system accounts" and "New user accounts"

The system MUST be rebooted after changing these settings and before installing Update 4.

After installing and testing Veeam I have changed the settings back and it appears to be working.

[Gostev]

Hi Jason, thank you so much for taking time to share the workaround with the community!

[mkretzer]

@gostev: Is it really a timezone thing or do we have to change locale for example from german to english?

[Moodie007]

I syspect the only thing you need to change is the system locale. Looking at the logs, it was trying to write a date to the database using the mm/dd/yyyy format. The system region was set to NZ. Using SQL SSMS I could see that it was using the NZ format of dd/mm/yyyy. So it was trying to write a date 01/30/2019. There are only 12 months not 30 hence the failure. When you change the system locale to US i could see in SQL SSMS that SQL is now using the American format.

Hope that makes sense and helps.

[DaveWatkins]

If that's the case this should be much more widespread I would have thought. USA is pretty much the only country that uses mm/dd/yyyy format as can be seen here

http://calendars.wikia.com/wiki/Date_format_by_country

[robcawood]

Hi All... I am also in NZ and have had the same error... here is the section from the log file...

Code: Select all

***Veeam***  SRV: [31.01.2019 11:39:43]:<00000001> [VEEAM] Database version has been incremented to <5186>
***Veeam***  SRV: [31.01.2019 11:39:43]:<00000001> [VEEAM] Executing version command (ContentType=[Schema], Version=<5186>)
***Veeam***  SRV: [31.01.2019 11:39:43]:<00000001> [VEEAM] Adding trusted hosts (InitialCatalog=[VeeamBackup], DatabaseVersion=<5186>)
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] declare @id uniqueidentifier;
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] set @id = '52107791-e0ff-4538-a26e-4717962aef77';
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] declare @usn bigint;
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] exec IncrementUsn @usn output;
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] BEGIN
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] INSERT INTO [dbo].[Backup.Model.HostSshFingerprint]
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] (
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] [id], [network_name], [accepted_fingerprint], [current_fingerprint], [public_key], [agent_id], [network_name_hash], [last_update_time], [usn]
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] )
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] VALUES  
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] (
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] @id, '99.99.99.99', '', '', '', '00000000-0000-0000-0000-000000000000', '', '31/01/2019 11:39:43 a.m.', @usn
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] )
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] END;
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM] [SQL|ERROR] GO
***Veeam***  SRV: [31.01.2019 11:39:44]:<00000001> [VEEAM|ERROR] Conversion failed when converting date and/or time from character string. (System.Data.SqlClient.SqlException)

Hope this helps resolve the issue... PS I have changed our IP address to 99.99.99.99

Regards

Rob

[robcawood]

Hi All... an update.

Changing the "System Locale" and rebooting didn't fix the issue for me, I had to change the Format (on the Regions Formats tab) to English (United States). This actually changed the date/time display to the US style 1/31/2019.

The update attempt after this change worked fine, here are the logs...

Code: Select all

***Veeam***  SRV: [31.01.2019 12:06:54]:<00000001> [VEEAM] Database version has been incremented to <5186>
***Veeam***  SRV: [31.01.2019 12:06:54]:<00000001> [VEEAM] Executing version command (ContentType=[Schema], Version=<5186>)
***Veeam***  SRV: [31.01.2019 12:06:54]:<00000001> [VEEAM] Adding trusted hosts (InitialCatalog=[VeeamBackup], DatabaseVersion=<5186>)
***Veeam***  SRV: [31.01.2019 12:06:54]:<00000001> [VEEAM] Trusted hosts was successfully added (InitialCatalog=[VeeamBackup])
***Veeam***  SRV: [31.01.2019 12:06:54]:<00000001> [VEEAM] [SQL|INFO] Adjusted reports cleared
***Veeam***  SRV: [31.01.2019 12:06:54]:<00000001> [VEEAM] Database version has been incremented to <5187>

Regards

Rob

[francisko]

Okay I have a workaround. I wasn’t through enough when I changed the region originally - I suspect I didn’t reboot after making the changes. You won't need to change all of these settings, no doubt - such as "New User Accounts" tick box, but this is what I did.

My Case ID: 03376372

Thanks for sharing this workaround.
I changed the following three settings, then I can successfully upgrade from U3A to U4.
Formats Tab - Format: English (United States)
Location Tab - Home Location: United States
Administrative Tab - Change system locale... English (United States)

So far, it looks work fine, and I'll keep watching it.

[Jooonas]

Same issue here, Same issue here, the workarounds mentioned above do not work in my environment.
I already openend a case (03390914).

[infused]

Can we seriously get a fix for this?

Support seems to know nothing. We've had to completely uninstall, re-install with update 4 and then restore the catalog (thats the support fix). It boils down to the installer doing something funny with the date in SQL if P.M is used instead of the US default.

It's a shit show.

(this also needs a sticky imo)

[Dima P.]

Hi folks,

This issue is being investigated by RnD team. I understand your frustration and really sorry for any inconvenience caused, we are doing our best to find the root cause. We will update this thread as soon as we have a confirmed solution.

Meanwhile please if you faced similar issue with Enterprise Manager upgrade please submit a case and share it with us. You can also check the workaround suggested by Moodie007. Thanks!

[mkretzer]

One question: If we run into this issue and want to try a workaround we can just re-try the upgrade without complete uninstall/reinstall?

[DaveWatkins]

Just to be clear, this isn't just EM that has this issue, and upgrade of B&R from U3A to U4 can also see it. That's when I saw it and as far as I can tell when a lot of others in this thread also saw it

[infused]

Correct. This is B&R for most people.

[engedib]

Relevant KB article with the workaround:
https://www.veeam.com/kb2884

[Gostev]

I've just updated the KB with the list of affected regional formats and countries.

[l.scotti]

French language is affected too.

[Dima P.]

Hello l.scotti,

Can you please share your case id for the reference? Thank you!

[l.scotti]

Hi Dima P,

I haven't opened a case yet but I will.
I'll share you case id

[l.scotti]

OK case id is 03420339

[myFist]

Hi folks,

This issue is being investigated by RnD team. I understand your frustration and really sorry for any inconvenience caused, we are doing our best to find the root cause. We will update this thread as soon as we have a confirmed solution.

Meanwhile please if you faced similar issue with Enterprise Manager upgrade please submit a case and share it with us. You can also check the workaround suggested by Moodie007. Thanks!

I faced this issue with the Enterprise Manager and Update 4a. The cause was the Service Account which I used (the same like I used for the B&R Installation). Installed with user Local System worked fine, changed the Service to my service account afterwards.