Since Sunday I have experienced "The semaphore timeout period has expired" errors when working with Azure Files shares e.g. when copying a 1.25 GB file. The problem is reproducible from all my VMs.

Hi MiLar,

This issue occurs if your network is very busy or looses the network connection.  For example, your network may be  very busy if many client machines are copying files to the server.

However to analyze the exact cause we might have to know the Error code of the exception like Error 50 etc., and need to understand the environment .

Regards,

Shirisha Paderu

• Proposed as answer by Shirisha PaderuMicrosoft contingent staff, Moderator 16 hours 40 minutes ago

Hi Shirisha

Thanks for replying.

The network is not busy. According to the task manager the network utilization is 0%.

The error is reproducible with only one VM started running Windows 8.1 (MSDN image).

The exact error message is (occurs about 30 minutes after starting to copy a 1.25 GB file from a mapped Azure Files share to the local C-drive. No progress at all is shown during the 30 minutes):

1 Interrupted Action

An unexpected error is keeping you from copying the file. If you continue to receive this error, you can use the error code to search for help with this problem.

Error 0x80070079: The semaphore timeout period has expired.

Try Again

Skip

Cancel

Regards,

Michael

Hi MiLar,

This issue occurs if your network is very busy or looses the network connection.  For example, your network may be  very busy if many client machines are copying files to the server.

However to analyze the exact cause we might have to know the Error code of the exception like Error 50 etc., and need to understand the environment .

Regards,

Shirisha Paderu

• Proposed as answer by Shirisha PaderuMicrosoft contingent staff, Moderator Tuesday, July 28, 2015 2:39 PM

Hi MiLar,

This issue occurs if your network is very busy or looses the network connection.  For example, your network may be  very busy if many client machines are copying files to the server.

However to analyze the exact cause we might have to know the Error code of the exception like Error 50 etc., and need to understand the environment .

Regards,

Shirisha Paderu

• Proposed as answer by Shirisha PaderuMicrosoft contingent staff, Moderator Tuesday, July 28, 2015 2:39 PM

Hi MiLar,

This issue occurs if your network is very busy or looses the network connection.  For example, your network may be  very busy if many client machines are copying files to the server.

However to analyze the exact cause we might have to know the Error code of the exception like Error 50 etc., and need to understand the environment .

Regards,

Shirisha Paderu

• Proposed as answer by Shirisha PaderuMicrosoft contingent staff, Moderator Tuesday, July 28, 2015 2:39 PM

Hi MiLar,

This issue occurs if your network is very busy or looses the network connection.  For example, your network may be  very busy if many client machines are copying files to the server.

However to analyze the exact cause we might have to know the Error code of the exception like Error 50 etc., and need to understand the environment .

Regards,

Shirisha Paderu

• Proposed as answer by Shirisha PaderuMicrosoft contingent staff, Moderator Tuesday, July 28, 2015 2:39 PM

Hi MiLar,

This issue occurs if your network is very busy or looses the network connection.  For example, your network may be  very busy if many client machines are copying files to the server.

However to analyze the exact cause we might have to know the Error code of the exception like Error 50 etc., and need to understand the environment .

Regards,

Shirisha Paderu

• Proposed as answer by Shirisha PaderuMicrosoft contingent staff, Moderator Tuesday, July 28, 2015 2:39 PM

Hi MiLar,

This issue occurs if your network is very busy or looses the network connection.  For example, your network may be  very busy if many client machines are copying files to the server.

However to analyze the exact cause we might have to know the Error code of the exception like Error 50 etc., and need to understand the environment .

Regards,

Shirisha Paderu

• Proposed as answer by Shirisha PaderuMicrosoft contingent staff, Moderator Tuesday, July 28, 2015 2:39 PM

Hi, 

When you receive semaphore timeout messages,
the messages are usually the result of one of the following

1. A heavy load on the server is causing processes to be delayed from releasing semaphores.

2. A process has crashed while holding a semaphore, causing other processes to block when trying to acquire the
semaphore.

3. Deadly embrace, semaphore contention where two tasks are waiting on each other and neither task is able to break the loop. In the simplest case, thread A is trying to get a semaphore which is owned by
B, while B is trying to get a different semaphore which is owned by A. More complex combinations are also possible: A wants a semaphore owned by B, who wants a semaphore owned by C, who wants a semaphore owned by A, etc.

4. If a process was to fail to set a semaphore during execution, another process dependent on the semaphore will be blocked awaiting the semaphore.

Please restart the Virtual Machine to see if the issue persists. Also disable the firewall to check the result.

Feel free to write back if that doesnt work.

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator 15 hours 36 minutes ago

Hi, 

When you receive semaphore timeout messages,
the messages are usually the result of one of the following

1. A heavy load on the server is causing processes to be delayed from releasing semaphores.

2. A process has crashed while holding a semaphore, causing other processes to block when trying to acquire the
semaphore.

3. Deadly embrace, semaphore contention where two tasks are waiting on each other and neither task is able to break the loop. In the simplest case, thread A is trying to get a semaphore which is owned by
B, while B is trying to get a different semaphore which is owned by A. More complex combinations are also possible: A wants a semaphore owned by B, who wants a semaphore owned by C, who wants a semaphore owned by A, etc.

4. If a process was to fail to set a semaphore during execution, another process dependent on the semaphore will be blocked awaiting the semaphore.

Please restart the Virtual Machine to see if the issue persists. Also disable the firewall to check the result.

Feel free to write back if that doesnt work.

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator Friday, July 31, 2015 3:42 PM

Hi, 

When you receive semaphore timeout messages,
the messages are usually the result of one of the following

1. A heavy load on the server is causing processes to be delayed from releasing semaphores.

2. A process has crashed while holding a semaphore, causing other processes to block when trying to acquire the
semaphore.

3. Deadly embrace, semaphore contention where two tasks are waiting on each other and neither task is able to break the loop. In the simplest case, thread A is trying to get a semaphore which is owned by
B, while B is trying to get a different semaphore which is owned by A. More complex combinations are also possible: A wants a semaphore owned by B, who wants a semaphore owned by C, who wants a semaphore owned by A, etc.

4. If a process was to fail to set a semaphore during execution, another process dependent on the semaphore will be blocked awaiting the semaphore.

Please restart the Virtual Machine to see if the issue persists. Also disable the firewall to check the result.

Feel free to write back if that doesnt work.

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator Friday, July 31, 2015 3:42 PM

Hi, 

When you receive semaphore timeout messages,
the messages are usually the result of one of the following

1. A heavy load on the server is causing processes to be delayed from releasing semaphores.

2. A process has crashed while holding a semaphore, causing other processes to block when trying to acquire the
semaphore.

3. Deadly embrace, semaphore contention where two tasks are waiting on each other and neither task is able to break the loop. In the simplest case, thread A is trying to get a semaphore which is owned by
B, while B is trying to get a different semaphore which is owned by A. More complex combinations are also possible: A wants a semaphore owned by B, who wants a semaphore owned by C, who wants a semaphore owned by A, etc.

4. If a process was to fail to set a semaphore during execution, another process dependent on the semaphore will be blocked awaiting the semaphore.

Please restart the Virtual Machine to see if the issue persists. Also disable the firewall to check the result.

Feel free to write back if that doesnt work.

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator Friday, July 31, 2015 3:42 PM

Hi, 

When you receive semaphore timeout messages,
the messages are usually the result of one of the following

1. A heavy load on the server is causing processes to be delayed from releasing semaphores.

2. A process has crashed while holding a semaphore, causing other processes to block when trying to acquire the
semaphore.

3. Deadly embrace, semaphore contention where two tasks are waiting on each other and neither task is able to break the loop. In the simplest case, thread A is trying to get a semaphore which is owned by
B, while B is trying to get a different semaphore which is owned by A. More complex combinations are also possible: A wants a semaphore owned by B, who wants a semaphore owned by C, who wants a semaphore owned by A, etc.

4. If a process was to fail to set a semaphore during execution, another process dependent on the semaphore will be blocked awaiting the semaphore.

Please restart the Virtual Machine to see if the issue persists. Also disable the firewall to check the result.

Feel free to write back if that doesnt work.

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator Friday, July 31, 2015 3:42 PM

Hi, 

When you receive semaphore timeout messages,
the messages are usually the result of one of the following

1. A heavy load on the server is causing processes to be delayed from releasing semaphores.

2. A process has crashed while holding a semaphore, causing other processes to block when trying to acquire the
semaphore.

3. Deadly embrace, semaphore contention where two tasks are waiting on each other and neither task is able to break the loop. In the simplest case, thread A is trying to get a semaphore which is owned by
B, while B is trying to get a different semaphore which is owned by A. More complex combinations are also possible: A wants a semaphore owned by B, who wants a semaphore owned by C, who wants a semaphore owned by A, etc.

4. If a process was to fail to set a semaphore during execution, another process dependent on the semaphore will be blocked awaiting the semaphore.

Please restart the Virtual Machine to see if the issue persists. Also disable the firewall to check the result.

Feel free to write back if that doesnt work.

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator Friday, July 31, 2015 3:42 PM

Hi, 

When you receive semaphore timeout messages,
the messages are usually the result of one of the following

1. A heavy load on the server is causing processes to be delayed from releasing semaphores.

2. A process has crashed while holding a semaphore, causing other processes to block when trying to acquire the
semaphore.

3. Deadly embrace, semaphore contention where two tasks are waiting on each other and neither task is able to break the loop. In the simplest case, thread A is trying to get a semaphore which is owned by
B, while B is trying to get a different semaphore which is owned by A. More complex combinations are also possible: A wants a semaphore owned by B, who wants a semaphore owned by C, who wants a semaphore owned by A, etc.

4. If a process was to fail to set a semaphore during execution, another process dependent on the semaphore will be blocked awaiting the semaphore.

Please restart the Virtual Machine to see if the issue persists. Also disable the firewall to check the result.

Feel free to write back if that doesnt work.

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator Friday, July 31, 2015 3:42 PM

Hi

I have restarted the VM several times and it is the only VM that access the Azure Files share.

I have disabled System Center Endpoint Protection and Windows Firewall.

I have found out that the problem only occurs with two specific ISO files on the share and disabling Windows Firewall and System Center Endpoint Protection do not solve the problem.

Regards,

Michael

Hi,

Were you able to copy the  files of same size other than ISO files ?

Would you be able to enable Azure file shares for any other VM and have a check if the copy succeeds ?

What is the size of the VM that has issues ? Could you increase the size of VM to the higher level and check if the issue persists.

Regards,

Shirisha Paderu

Hi

Thanks for following up on this issue.

Yes, I can copy other ISO files of the same size or even larger from the same share without problems.

The problem can be reproduced on another VM using the same Azure Files share and ISO files.

I have changed the size from A3 to A4, but that does not solve the problem.

It seems that Azure Files has some kind of lock on these files making them inaccessible.

Regards,

Michael

Hi,

We are checking on this and would get back to you on our findings.

Regards,

ShirshaPaderu

Hi,

Regarding this case, we would like to have some more details as below:

-          What is the storage account/share/directory/file that has the issue? Please list all files that got the issue.

-          What is the operating system of the VM mounting the file share and what is the VM name? Please list all the VMs.

-          How are you  trying to upload/download the ISO file? With powershell, AzCopy or directly coping on the mounted drive?

-          Is the issue intermittent or still occurring?

-          Is it possible that you capture a network trace with tools such as Netmon or Wireshark when re-producing this issue?

Regards,

Shirisha Paderu

Hi

Is it possible to send the storage account information etc. as private information?

Regards,

Michael

Hi,

You may send it to my email " v-shpade at Microsoft dot com "

Regards,

Shirisha Paderu

Hi,

We have engaged our senior engineers tolook into the issue :

They would like to know if the  issue occurs on all VMs? and If the issue occurs for all ISO files?

Have you tested other ISO files in addition to the two mentioned?

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator 20 hours 9 minutes ago

Hi

I only have 2 VMs, and the problem occurs on both.

The issue does not occur for all ISO files. I have tested with other ISO files of the same size or larger and on the same share and they work.

Regards,

Michael

Hi,

We have engaged our senior engineers tolook into the issue :

They would like to know if the  issue occurs on all VMs? and If the issue occurs for all ISO files?

Have you tested other ISO files in addition to the two mentioned?

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator Friday, August 14, 2015 11:09 AM

Hi,

We have engaged our senior engineers tolook into the issue :

They would like to know if the  issue occurs on all VMs? and If the issue occurs for all ISO files?

Have you tested other ISO files in addition to the two mentioned?

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator Friday, August 14, 2015 11:09 AM

Hi,

We have engaged our senior engineers tolook into the issue :

They would like to know if the  issue occurs on all VMs? and If the issue occurs for all ISO files?

Have you tested other ISO files in addition to the two mentioned?

Regards,

Shirisha Paderu

• Edited by Shirisha PaderuMicrosoft contingent staff, Moderator Friday, August 14, 2015 11:09 AM

Hi Mila97,

I will be taking over this thread from now. We will update you from our seniors at the earliest.

Girish Prajwal

Hi

Do you have an update on this problem?

Regards,

Michael