Site Manager V7.2.4841 & Reflect V7.2.4808 Intermittent Performance Issues and Intermitat Repository...


Site Manager V7.2.4841 & Reflect V7.2.4808 Intermittent Performance...
Author
Message
Morris
Morris
Junior Member
Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)
Group: Forum Members
Posts: 58, Visits: 95
The products seem to be conflicting with each other.  

My backups are all run via Site Manager.  Yesterday most of my Windows 10 Workstations were upgraded to Reflect V7.2.4808.  Incremental backups ran normally.  

Today I upgraded the Windows 10 computer that runs Site Manager.  I started getting access errors to to the Repositories which are windows shares residing on the Site Manager computer.  When I visited the Site Manager computer I noticed that it had a Reflect update pending and I did the update. My incromentals started running yet taking much longer than usual.  I restarted the server and started getting access issues on some of the workstations.  I restarted the workstations and then slow incromentals on some computers.  The problems move back and forth between these two states depending on which computers are restarted last.  

I have one workstation with a huge dynamic disk and the incromental of the partition is estimated to take 7 hours.  It used to take minutes.  The large dynamic partition is 18TB and 8.95TB are used.

Please assist.
Thank  you,
Morris
Morris
Morris
Junior Member
Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)
Group: Forum Members
Posts: 58, Visits: 95
Morris - 27 March 2020 11:06 PM
The products seem to be conflicting with each other.  

My backups are all run via Site Manager.  Yesterday most of my Windows 10 Workstations were upgraded to Reflect V7.2.4808.  Incremental backups ran normally.  

Today I upgraded the Windows 10 computer that runs Site Manager.  I started getting access errors to to the Repositories which are windows shares residing on the Site Manager computer.  When I visited the Site Manager computer I noticed that it had a Reflect update pending and I did the update. My incromentals started running yet taking much longer than usual.  I restarted the server and started getting access issues on some of the workstations.  I restarted the workstations and then slow incromentals on some computers.  The problems move back and forth between these two states depending on which computers are restarted last.  

I have one workstation with a huge dynamic disk and the incromental of the partition is estimated to take 7 hours.  It used to take minutes.  The large dynamic partition is 18TB and 8.95TB are used.

Please assist.
Thank  you,
Morris

Looking at Task Manager there is very heavy disk access.  I see the same level of access during a full backup.  It feels to me like the entire disk is being read even though there is no network IO to speak of.

Morris
Morris
Junior Member
Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)
Group: Forum Members
Posts: 58, Visits: 95
The Incremental completed and the backup file size is 800GB.  Should have been about 50GB.  
Morris
Morris
Junior Member
Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)
Group: Forum Members
Posts: 58, Visits: 95
Tried to run the incremental on the large partition again.  No computers were restarted though the workstation had slept after the previous incrometal and had this access issue:

Destination:
Backup Type:    Full - Incremental specified but no image set to append to
File Name:    \\srv\Big3\MORRIS-PC\{BA2B4002-68C6-4959-A05D-14D49FC77BE2}\12D28C5317FA8E90-00-00.mrimg
Attempting to connect to: '\\srv\Big3\MORRIS-PC\{BA2B4002-68C6-4959-A05D-14D49FC77BE2}'
Failure: User - No user - Access is denied.
Failure: User - anonymous - The user name or password is incorrect.

It is one of two type of access issues I've seen.  The other is:

Backup Type:    Incremental
File Name:    Append to recent image in directory '\\srv\backup\GERALD-PC\{43BB4DC6-5CD0-4433-A383-8A591CFCC064}\'
Attempting to connect to: '\\srv\backup\GERALD-PC\{43BB4DC6-5CD0-4433-A383-8A591CFCC064}'
Failure: User - morris - Access is denied.
Failure: User - No user - Access is denied.
Failure: User - anonymous - The user name or password is incorrect.
\\srv\backup\GERALD-PC\{43BB4DC6-5CD0-4433-A383-8A591CFCC064}\97BCC5983FB1B575-05-05.mrimg

Morris is the user logged into the Site Manager Site.  The partition is only accessible by user Backup and the configuration for the backup shows Backup as the user.
Alex
Alex
Macrium Representative
Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)
Group: Moderators
Posts: 275, Visits: 941
Hi Morris,

Firstly, apologies for the problems you're encountering, I have a few general comments which might be helpful, please let me know I'm misunderstanding anything:

1. Repository access - we've noticed that in recent months of Windows 10 updates have caused some share access issues, it may be worth checking to see when the last updates were applied to see if this might be causing any problems. According to a few internet sources, these sometimes can be resolved by changing the repository to use IP address rather than server name in the share UNC path.

2. On the agent, there are some logs in c:\programdata\macrium\sitemanager\agentnetwork.log which contain the Win32 error codes for the network access attempts, these may help identify the problem.

3. With regard to the backup size and time, Site Manager 4814 includes all changes from the latest Reflect release - http://updates.macrium.com/reflect/v7/v7.2.4808/details7.2.4808.htm . This includes the additional checks and features for data integrity mentioned in the 7.2.4797 release section - the first incremental taken in any backup set with the new Site Manager release will take longer and involve more disk IO to perform additional integrity checks against the source disks. 

That said, a size disparity in incrementals like the one you're seeing is unusual and we'd like to investigate in more detail. We should also be able to provide some assistance with what's going on with the repository share access problems.
Can I ask that you open a support case via https://helpme.macrium.com/ and include a link to this forum post? We can then get the process of collecting logs and and reviewing them started and hopefully get things sorted out for you.

Kind Regards,

Alex

Macrium Development

Next Webinar

See our reviews on

Trustpilot Logo
Trustpilot Stars


Morris
Morris
Junior Member
Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)
Group: Forum Members
Posts: 58, Visits: 95
Alex - 28 March 2020 1:52 PM
Hi Morris,

Firstly, apologies for the problems you're encountering, I have a few general comments which might be helpful, please let me know I'm misunderstanding anything:

1. Repository access - we've noticed that in recent months of Windows 10 updates have caused some share access issues, it may be worth checking to see when the last updates were applied to see if this might be causing any problems. According to a few internet sources, these sometimes can be resolved by changing the repository to use IP address rather than server name in the share UNC path.

2. On the agent, there are some logs in c:\programdata\macrium\sitemanager\agentnetwork.log which contain the Win32 error codes for the network access attempts, these may help identify the problem.

3. With regard to the backup size and time, Site Manager 4814 includes all changes from the latest Reflect release - http://updates.macrium.com/reflect/v7/v7.2.4808/details7.2.4808.htm . This includes the additional checks and features for data integrity mentioned in the 7.2.4797 release section - the first incremental taken in any backup set with the new Site Manager release will take longer and involve more disk IO to perform additional integrity checks against the source disks. 

That said, a size disparity in incrementals like the one you're seeing is unusual and we'd like to investigate in more detail. We should also be able to provide some assistance with what's going on with the repository share access problems.
Can I ask that you open a support case via https://helpme.macrium.com/ and include a link to this forum post? We can then get the process of collecting logs and and reviewing them started and hopefully get things sorted out for you.

Hi Alex,

I recall reading about networking issues related to a Windows patch.  All computers applied the latest updates there about March 12 and I've had no problems till I updated Site Manager.  This makes me believe it's site manager.  Further more, the recent patches had nothing to do with networking, or at least what was listed.

Using the IP verses the server name has no effect.

I've done a lot of testing since trying this.  My findings:
- Multiple access to the same repository seems to cause the share access violations, on all or any but the first computer to start the incremental.  This behaviour seem to be consistent.
- When running one incremental at a time after the error is induced by the above the failure happens for a while and then I'm able to run incrementals consistently on the workstation that functions.  The length of time required till an individual workstation starts to work varies.  One may start working while another continues to fail for a while.  The failing one eventually starts to work.

Temporary action I'm taking:
-I'm setting the repository to limit one backup at a time
-I'm reviewing the retry on error settings.  I believe I turned this off because of a previous issue

I'll report back on the results of the above

I'll look at the large partition incremental and it's behaviour once I stabilize my other incromentals

Morris
Morris
Junior Member
Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)
Group: Forum Members
Posts: 58, Visits: 95
Morris - 28 March 2020 3:59 PM
Alex - 28 March 2020 1:52 PM
Hi Morris,

Firstly, apologies for the problems you're encountering, I have a few general comments which might be helpful, please let me know I'm misunderstanding anything:

1. Repository access - we've noticed that in recent months of Windows 10 updates have caused some share access issues, it may be worth checking to see when the last updates were applied to see if this might be causing any problems. According to a few internet sources, these sometimes can be resolved by changing the repository to use IP address rather than server name in the share UNC path.

2. On the agent, there are some logs in c:\programdata\macrium\sitemanager\agentnetwork.log which contain the Win32 error codes for the network access attempts, these may help identify the problem.

3. With regard to the backup size and time, Site Manager 4814 includes all changes from the latest Reflect release - http://updates.macrium.com/reflect/v7/v7.2.4808/details7.2.4808.htm . This includes the additional checks and features for data integrity mentioned in the 7.2.4797 release section - the first incremental taken in any backup set with the new Site Manager release will take longer and involve more disk IO to perform additional integrity checks against the source disks. 

That said, a size disparity in incrementals like the one you're seeing is unusual and we'd like to investigate in more detail. We should also be able to provide some assistance with what's going on with the repository share access problems.
Can I ask that you open a support case via https://helpme.macrium.com/ and include a link to this forum post? We can then get the process of collecting logs and and reviewing them started and hopefully get things sorted out for you.

Hi Alex,

I recall reading about networking issues related to a Windows patch.  All computers applied the latest updates there about March 12 and I've had no problems till I updated Site Manager.  This makes me believe it's site manager.  Further more, the recent patches had nothing to do with networking, or at least what was listed.

Using the IP verses the server name has no effect.

I've done a lot of testing since trying this.  My findings:
- Multiple access to the same repository seems to cause the share access violations, on all or any but the first computer to start the incremental.  This behaviour seem to be consistent.
- When running one incremental at a time after the error is induced by the above the failure happens for a while and then I'm able to run incrementals consistently on the workstation that functions.  The length of time required till an individual workstation starts to work varies.  One may start working while another continues to fail for a while.  The failing one eventually starts to work.

Temporary action I'm taking:
-I'm setting the repository to limit one backup at a time
-I'm reviewing the retry on error settings.  I believe I turned this off because of a previous issue

I'll report back on the results of the above

I'll look at the large partition incremental and it's behaviour once I stabilize my other incromentals

I can't say for certain that simultaneous backups causes the problem as I've seen workstations work and not work randomly.  One thing is clear and that is if a workstation backup works, it usually works multiple times as if one had primed a pump.   Then for an unknown reason that workstation backup will start having the authorization issue again and after some time it begins to work again.

I've been triggering the backups manually.  The retry timers are for scheduled jobs.  

Do you know the KB of the update that might be causing the access issue?
Morris
Morris
Junior Member
Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)
Group: Forum Members
Posts: 58, Visits: 95
The incromental of the large partition now runs quickly.  The problem was related to the changes.   This backup runs to a different repository and no other computers access it.  I've had the intermittent access issue to this repository as well as the one used for workstation incrementals.  The access issue seems rather random.

I believe the issue is with Site Manager and not windows SMB file sharing.  I'm done with my testing till I hear back from you.
Morris
Morris
Junior Member
Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)Junior Member (81 reputation)
Group: Forum Members
Posts: 58, Visits: 95
All of my scheduled backups ran normally overnight and also this afternoon.  I still had the workstation backups run 1 at a time rather than 3 simultaneously.  I'm leaving it as is for at least one more backup cycle.  At this point I no longer have a hunch if Sige Manager or Windows Networking is causing the intermittent access issue.  

Stay well,

Morris
Alex
Alex
Macrium Representative
Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)Macrium Representative (515 reputation)
Group: Moderators
Posts: 275, Visits: 941
Morris - 28 March 2020 4:42 PM
Morris - 28 March 2020 3:59 PM
Alex - 28 March 2020 1:52 PM
Hi Morris,

Firstly, apologies for the problems you're encountering, I have a few general comments which might be helpful, please let me know I'm misunderstanding anything:

1. Repository access - we've noticed that in recent months of Windows 10 updates have caused some share access issues, it may be worth checking to see when the last updates were applied to see if this might be causing any problems. According to a few internet sources, these sometimes can be resolved by changing the repository to use IP address rather than server name in the share UNC path.

2. On the agent, there are some logs in c:\programdata\macrium\sitemanager\agentnetwork.log which contain the Win32 error codes for the network access attempts, these may help identify the problem.

3. With regard to the backup size and time, Site Manager 4814 includes all changes from the latest Reflect release - http://updates.macrium.com/reflect/v7/v7.2.4808/details7.2.4808.htm . This includes the additional checks and features for data integrity mentioned in the 7.2.4797 release section - the first incremental taken in any backup set with the new Site Manager release will take longer and involve more disk IO to perform additional integrity checks against the source disks. 

That said, a size disparity in incrementals like the one you're seeing is unusual and we'd like to investigate in more detail. We should also be able to provide some assistance with what's going on with the repository share access problems.
Can I ask that you open a support case via https://helpme.macrium.com/ and include a link to this forum post? We can then get the process of collecting logs and and reviewing them started and hopefully get things sorted out for you.

Hi Alex,

I recall reading about networking issues related to a Windows patch.  All computers applied the latest updates there about March 12 and I've had no problems till I updated Site Manager.  This makes me believe it's site manager.  Further more, the recent patches had nothing to do with networking, or at least what was listed.

Using the IP verses the server name has no effect.

I've done a lot of testing since trying this.  My findings:
- Multiple access to the same repository seems to cause the share access violations, on all or any but the first computer to start the incremental.  This behaviour seem to be consistent.
- When running one incremental at a time after the error is induced by the above the failure happens for a while and then I'm able to run incrementals consistently on the workstation that functions.  The length of time required till an individual workstation starts to work varies.  One may start working while another continues to fail for a while.  The failing one eventually starts to work.

Temporary action I'm taking:
-I'm setting the repository to limit one backup at a time
-I'm reviewing the retry on error settings.  I believe I turned this off because of a previous issue

I'll report back on the results of the above

I'll look at the large partition incremental and it's behaviour once I stabilize my other incromentals

I can't say for certain that simultaneous backups causes the problem as I've seen workstations work and not work randomly.  One thing is clear and that is if a workstation backup works, it usually works multiple times as if one had primed a pump.   Then for an unknown reason that workstation backup will start having the authorization issue again and after some time it begins to work again.

I've been triggering the backups manually.  The retry timers are for scheduled jobs.  

Do you know the KB of the update that might be causing the access issue?

Hi Morris,

Unfortunately I don't have a KB - the information on the internet regarding the SMB issues is a little conflicting and confused, I think the best route is to get our support team to have a look at the detailed error logs and that should give us more of a clue as to what may be causing the issue.

I believe there were no significant changes to how Site Manager and Reflect interacts with SMB in the newest release - there were some code changes to better clean up and log the accesses. If it looks like this is causing a problem (the agentNetwork.log and agentProcess.log files on the agent will contain more information on this) we could look at providing a build which disables these features.

Kind Regards,

Alex

Macrium Development

Next Webinar

See our reviews on

Trustpilot Logo
Trustpilot Stars


GO

Merge Selected

Merge into selected topic...



Merge into merge target...



Merge into a specific topic ID...




Reading This Topic

Login

Explore
Messages
Mentions
Search