Skip to main content

Windows 10 Anniversary Update - some issues

Thread needs solution

Hey there,

last weekend I updated all workstations from Windows 10 1511 to 1607 (the anniversary update).
All servicepacks are applied.

In the night from sunday to monday I ran the backup plans to backup the workstations.
Each workstation
- has the agent installed
- is waked by WOL

The backup plans
- do a full backup to a managed vault on a NAS
- replicate the backup to a second managed vault on another NAS
- should send a mail (default options)
- run a shutdown script after backup
 

On monday none of the jobs ran successfully and all workstations were still up and running.

The error messages were quite weird, some were reporting that the user cancelled the job (there was no person present...), others reported that the status must be repaired....the status in the backup plan view on the AMS was also messed up.
If I restarted it, the status of the jobs changed from "OK" to "Error" and back again....

As I read in the forum, Windows 1607 had some problems with drive order....so I opened each plan and  removed all backup sources and configured it new.
The rest of the plan was not changed.
 

This night the backups ran properly....
But no mail...hmm

I think best would be a full reinstall of the agents (also using the celanup tool).
But I´m unsure if this can be done easily without loosing the settings.
The backup plans are all central ones on the AMS.
So I think it should be possible without loosing any plan settings or is any manual stuff necessary?

Thanks for any hints

Sven

 

0 Users found this helpful

Hello Sven,

I'm not sure I would recommend an agent reinstall in this case. However, if you decide to do this, you won't lose the plans, but you will have readd the machines to your AMS and reapply those plans (and also possibly groups etc.)

What happens with a full reinstall is this: the internal GUID that uniquely identifies the agent is removed from the registry duing cleanup. A new installation creates a new GUID for the agent. When you add this agent to the AMS, it will see it as a different machine because the GUID will be different.

So all your old agents will be grayed out in the AMS after clean-up and reappear after you add them.

After this, you will have to edit the exitsting plans to reapply them to the readded agents.

Hello Igor,

I did a reinstall and modified the backup plan.

First try : manual
I removed the "command after backup" setting and all worked fine.....

 

Second try : scheduled
I added the "command after backup" again.
It´s a batch with the command "shutdown /s /t 300 /c "shutdwon by Arconis" "

I gave it a try at scheduled runtime this night at 2:00AM --> no success.....huh ?!?!

The workstation was shutdown so the batch ran properly.
I looked into the AMS....in the backup plan view the plan was shown running and stuck at 71% (replicating to 1. location).
I clicked in the topmost corner "running actions"....nothing....huh ?!?!

Clicking "details" in the job itself showed me the running actions and also the log.
The log stated that it was stuck at replicating....

So I restarted the workstation and then it was even more strange:
The backup failed with some strange messages in the log (I´m currently not at the machine or AMS so I cannot copy them in this thread but one was something with shutting down or ending a service or application) and the status "error".
It seems as if this strange behaviour and the failure has something to do with the command after backup and/or the handshake between the agent and the AMS ?!?

What I´m wondering:
a) Mail
I setup the backup plan and also the AMS to send email if there´s an error....but I got no mail.

b) command after backup
As mentioned above I updated the machines from Windows 10 1511 to 1607 (anniversary update).
In the same task I also changed the command in the batch command from

shutdown /s /t 10 /c "shutdown by Arconis" --> shutdown /s /t 300 /c "shutdown by Arconis" "

The only thing that was changed was the time....

I will give it a try this evening with the old 10 seconds...

 

One thing I´m also wondering:
In the log of the successful backup plan I see some log entries (a bit shortened):
- backup of disk --> okay this is the backup of the agent to the vault on the first NAS
- replication/cleanup 1st location --> I setup to keep the last 4 weeks of backup but why is data transferred to vault ont he second NAS
- replication/cleanup 1st location --> again ? okay maybe this is the real cleanup task
- replication/cleanup 2nd location --> hmm what´s that?
- replication/cleanup 2nd location --> again ? cleanup of the "what´s that" ?

Is there any explanation?

Regards

Sven

 

 

 

 

 

 

 

Hi @all.

I did some tests this evening with the "command after backup".

The tests were done on 2 different machines.
One (NB1) that was only updated from W10 1511 to 1607, and the second one (PC3) where I also uninstalled the Acronis Tools from the workstations and reinstalled them. 

a) Scheduled task, no "command after backup"
No problems on both machines!

After the succesful plan execution I analyzed the logs and saw one interesting point:
The catalog is updated right after the first task "backup" is finished (I setup the plans to do a full catalog update).
The replication and data retention for the first location is started with a delay of about 1 minute after the backup is finished.
A second catalog update is started right after teh replication command is done.
There is also a delay of about 30 seconds between finishing the replication to 1sr location and the start of replication to 2nd location.
And then I receive the mail of the successful execution.

 

b) Scheduled backup WITH "command after backup" and 3 seconds delay to shutdown
On PC3 I setup the plan to run a BATCH file after backup with the command

shutdown /s /t 3 /c "Shutdown by ACRONIS"

I look into AMS and the status of the backup plan shows 71% completed.
The only entry is "backup of volume" and nothing more.
I restarted the workstation and the AMS showed a message box with the information of the failed backup plan, again no mail was sent.

The result in the log:

The backup command is executed, the log states the successful execution.
The next entry has a timestamp of about 30 seconds later:

ProtectionAggregation: Protection object update error, some information may be incorrect.
Zusätzliche Info:
--------------------
Fehlercode: 1
Module: 307
Zeileninfo: ec69ad9eeeb1a08e
Felder:  $module : agent_protection_addon_vs_44421
Nachricht: ProtectionAggregation: Protection object update error, some information may be incorrect.
--------------------
Fehlercode: 10
Module: 249
Zeileninfo: 6298a2d9c3a1e7c3
Felder:  $module : mms_vs_44421
Nachricht: Der Dienst wird gerade gestoppt.
--------------------

 

 

c) Scheduled backup WITH "command after backup" and 300 seconds delay to shutdown
On NB1 I setup the plan to run a BATCH file after backup with the command

shutdown /s /t 300 /c "Shutdown by ACRONIS"

Again I took a look into AMS and the status of the backup plan shows 71% completed.

Again I restarted the workstation and the AMS showed a message box with the information of the failed backup plan, again no mail was sent.

The entries in the log show the execution of backup action.
I also see that the agent requested the replication/retention to be executed by the Storage Node Machine with the upper mentioned delay of about 1 minute.
And even the start of replication is shown in the log (executed by the storage node machine).

But then - suprise - after aboput 5 minutes after the backup task completed the following error messages are in the log

ASYNC: Full barrier action job.
Zusätzliche Info:
--------------------
Fehlercode: 4
Module: 249
Zeileninfo: 368cb7423c1c08d9
Felder:  TraceLevel : 1, $module : mms_vs_44421
Nachricht: ASYNC: Full barrier action job.
--------------------
Fehlercode: 4
Module: 249
Zeileninfo: d7761d47edeb39a1
Felder:  TraceLevel : 1, $module : mms_vs_44421
Nachricht: ASYNC: Action sequence pending action job.
--------------------
Fehlercode: 4
Module: 249
Zeileninfo: d0300d0ab9c42eb0
Felder:  TraceLevel : 1, $module : mms_vs_44421
Nachricht: Undefined job
--------------------
Fehlercode: 4
Module: 249
Zeileninfo: 40b6b675b5f23512
Felder:  TraceLevel : 1, $module : mms_vs_44421
Nachricht: ASYNC: Data binding job.
--------------------
Fehlercode: 10
Module: 249
Zeileninfo: 6298a2d9c3a1e7c3
Felder:  $module : mms_vs_44421
Nachricht: Der Dienst wird gerade gestoppt.
--------------------

 

Thie log enrtry occurs 3 times !!

 

From my point of view the "command after backup" is excuted right after the backup task (yes this is stated also in the manual).

BUT 

Why was the backup successful with Windows 10 1511 before the anniversary update and a delay of 10 seconds in the shutdown command?

Why is no e-mail sent if the packup plan fails ?!?

 

 

It seems as if the delay in the shutdown command is extremely critical.

So is there any idea to shutdown a workstation with an agent after the backup plan when I want replication ?

 

I´m wondering why Acronis has this feature of replication and validation but there is no way to shutdown the machines.
 

 

Regards

Sven