We were previously running VisualCron version 8.3.6 and upgraded to version 8.5.4.

We have a number of jobs which use the VisualCron "write file" task to write to a file on a network share (its a NetApp Network Attached Storage Server, presenting windows shares) . The file is a text file.

Periodically the write task will fail with an exit code of 77777, and the stderr output is recording the error message, "Exception in Task: Invalid Signature". When the error occurs, rerunning the failed task works (no error generated and the task completes successfully). This problem was not seen on version 8.3.6 for us, and only materialized after we upgraded to 8.5.4.

We upgraded to BETA version 8.5.5 and have been running it for about a day, the problem has not resurfaced. Is this a known problem corrected in 8.5.5?

Hi,

I spoke too soon, error just reoccurred. Same VisualCron task, Write File, but different job.

Hi,

I spoke too soon, error just reoccurred. Same VisualCron task, Write File, but different job.

Hi,

Could you take a screenshot of all settings in the problematic task and send it to support@visualcron.com, please?

---

Hi,

I spoke too soon, error just reoccurred. Same VisualCron task, Write File, but different job.

Hi,

Could you take a screenshot of all settings in the problematic task and send it to support@visualcron.com, please?

Hi,

I've uploaded screenshot of the write file task settings.  invalidSignature.docx (256kb) downloaded 45 time(s).

Hi,

I spoke too soon, error just reoccurred. Same VisualCron task, Write File, but different job.

Hi,

Could you take a screenshot of all settings in the problematic task and send it to support@visualcron.com, please?

Hi,

I've uploaded screenshot of the write file task settings.  invalidSignature.docx (256kb) downloaded 45 time(s).

Thank you. Now, could you try the task with those settings but locally instead and no credential? Trying to break down the issue to see if it's credential/network-related.

---

Hi,
I created a new job (with multiple Write file tasks) on my local C: drive. The Write file task is writing to my local C: drive (no credentials required). I ran the job once ever minute, for about 8 hours. I could not reproduce the error.

I did some further analysis of the machines with the errors. These machines have anywhere from 150 to 250 jobs scheduled with VisualCron. On these machines we have 2 special jobs.

1. Job #1 - checks for jobs which have been inactivated, activated or deleted. If any of one these 3 things happen, an email is sent and a file is written to a network share. The trigger used is "Event - VisualCron", job Job-inactivated, Job-activated, Job-deleted). Polling for the trigger is set at 900 seconds

2. Job #2 - checks for a job which has a failed task but completes. If this occurs an email is sent and a file is written to a network share. The trigger used for this job is "Event - VisualCron, Job completed -failed, ". Polling for the trigger is set for 60 seconds

I changed the polling to 1800 seconds for both of these 2 special jobs. And the machines have been running without errors (ie "clean") for almost 24 hours.
Some more clarity on the errors seen with a higher polling frequency (60 to 900 seconds) are, "invalid signature", "file not found", "...attached resource not found..."

Questions:
1. What is the recommended polling setting?
2. What is being polled (example, a log file? or each of the jobs?)
3. When I set the polling value to 1800 seconds, does this mean every job (?) or a log file (?) is checked every 1800 seconds?

Polling? Are you talking about the File Trigger? Then you should not use polling unless network server is a Samba share (Linux). In WIndows polling is not used and reaction is "instant".

---

Polling? Are you talking about the File Trigger? Then you should not use polling unless network server is a Samba share (Linux). In WIndows polling is not used and reaction is "instant".

Hi,
I just upload screen shot for the event trigger "polling" setting.



Definitely the write file tasks and read file and even the list file tasks we use (which seem to fail) correspond to a share (our windows boxes see them as shares, drive letter or a UNC path). The share physically correspond to a NAS (network attached storage server). It NAS (controller is a Linux box), presenting a CIF share.

What does the polling setting do?

Polling? Are you talking about the File Trigger? Then you should not use polling unless network server is a Samba share (Linux). In WIndows polling is not used and reaction is "instant".

Hi,
I just upload screen shot for the event trigger "polling" setting.



Definitely the write file tasks and read file and even the list file tasks we use (which seem to fail) correspond to a share (our windows boxes see them as shares, drive letter or a UNC path). The share physically correspond to a NAS (network attached storage server). It NAS (controller is a Linux box), presenting a CIF share.

What does the polling setting do?

I'm sorry. I think I'm making the problem description confusing.
Problem
- Jobs with tasks (Write file, List file, Read file), fail seemingly at random
- These jobs reside on multiple machines. Some machines have close to 250 VisualCron jobs

As an attempt to alert us of a failed task, we configured a job to check the outcome of the failed job. Example, if the job completed but had a failed task, send an email message and then write a file. We used the event trigger, VisualCron (see the previous screen attachements)

Likewise we also have a job which checks for jobs which have been inactivated, activated or deleted.

Think of these two jobs as "Checking type jobs". These two jobs are not failing per se. But we see a pattern where we believe these checking jobs are making the other jobs fail.

Polling? Are you talking about the File Trigger? Then you should not use polling unless network server is a Samba share (Linux). In WIndows polling is not used and reaction is "instant".

Hi,
I just upload screen shot for the event trigger "polling" setting.



Definitely the write file tasks and read file and even the list file tasks we use (which seem to fail) correspond to a share (our windows boxes see them as shares, drive letter or a UNC path). The share physically correspond to a NAS (network attached storage server). It NAS (controller is a Linux box), presenting a CIF share.

What does the polling setting do?

I'm sorry. I think I'm making the problem description confusing.
Problem
- Jobs with tasks (Write file, List file, Read file), fail seemingly at random
- These jobs reside on multiple machines. Some machines have close to 250 VisualCron jobs

As an attempt to alert us of a failed task, we configured a job to check the outcome of the failed job. Example, if the job completed but had a failed task, send an email message and then write a file. We used the event trigger, VisualCron (see the previous screen attachements)

Likewise we also have a job which checks for jobs which have been inactivated, activated or deleted.

Think of these two jobs as "Checking type jobs". These two jobs are not failing per se. But we see a pattern where we believe these checking jobs are making the other jobs fail.

And the number occurrences of jobs with a failed task seem to increase as we decrease the polling frequency setting. Originally the checker type jobs, event trigger polling setting was set to 60 and 900 seconds. We increased the polling to 1800 seconds for both checker type jobs, and the environment has been running clean (no job failures) for about 24 hours.

We see a correlation between this polling setting and jobs (task failing).

Polling? Are you talking about the File Trigger? Then you should not use polling unless network server is a Samba share (Linux). In WIndows polling is not used and reaction is "instant".

Hi,
I just upload screen shot for the event trigger "polling" setting.



Definitely the write file tasks and read file and even the list file tasks we use (which seem to fail) correspond to a share (our windows boxes see them as shares, drive letter or a UNC path). The share physically correspond to a NAS (network attached storage server). It NAS (controller is a Linux box), presenting a CIF share.

What does the polling setting do?

I'm sorry. I think I'm making the problem description confusing.
Problem
- Jobs with tasks (Write file, List file, Read file), fail seemingly at random
- These jobs reside on multiple machines. Some machines have close to 250 VisualCron jobs

As an attempt to alert us of a failed task, we configured a job to check the outcome of the failed job. Example, if the job completed but had a failed task, send an email message and then write a file. We used the event trigger, VisualCron (see the previous screen attachements)

Likewise we also have a job which checks for jobs which have been inactivated, activated or deleted.

Think of these two jobs as "Checking type jobs". These two jobs are not failing per se. But we see a pattern where we believe these checking jobs are making the other jobs fail.

And the number occurrences of jobs with a failed task seem to increase as we decrease the polling frequency setting. Originally the checker type jobs, event trigger polling setting was set to 60 and 900 seconds. We increased the polling to 1800 seconds for both checker type jobs, and the environment has been running clean (no job failures) for about 24 hours.

We see a correlation between this polling setting and jobs (task failing).

Thanks for the information, we'll take a closer look at this. Do you see any correlation between the trigger you are using for the Write File, List File, Read File job/tasks? What trigger is being used for them?

---

Polling? Are you talking about the File Trigger? Then you should not use polling unless network server is a Samba share (Linux). In WIndows polling is not used and reaction is "instant".

Hi,
I just upload screen shot for the event trigger "polling" setting.



Definitely the write file tasks and read file and even the list file tasks we use (which seem to fail) correspond to a share (our windows boxes see them as shares, drive letter or a UNC path). The share physically correspond to a NAS (network attached storage server). It NAS (controller is a Linux box), presenting a CIF share.

What does the polling setting do?

I'm sorry. I think I'm making the problem description confusing.
Problem
- Jobs with tasks (Write file, List file, Read file), fail seemingly at random
- These jobs reside on multiple machines. Some machines have close to 250 VisualCron jobs

As an attempt to alert us of a failed task, we configured a job to check the outcome of the failed job. Example, if the job completed but had a failed task, send an email message and then write a file. We used the event trigger, VisualCron (see the previous screen attachements)

Likewise we also have a job which checks for jobs which have been inactivated, activated or deleted.

Think of these two jobs as "Checking type jobs". These two jobs are not failing per se. But we see a pattern where we believe these checking jobs are making the other jobs fail.

And the number occurrences of jobs with a failed task seem to increase as we decrease the polling frequency setting. Originally the checker type jobs, event trigger polling setting was set to 60 and 900 seconds. We increased the polling to 1800 seconds for both checker type jobs, and the environment has been running clean (no job failures) for about 24 hours.

We see a correlation between this polling setting and jobs (task failing).

Thanks for the information, we'll take a closer look at this. Do you see any correlation between the trigger you are using for the Write File, List File, Read File job/tasks? What trigger is being used for them?

Hi,
No I don't see a correlation with that trigger. The jobs with the failing tasks (write file, list file, read file) are triggered by the very common, CUSTOM TIME TRIGGER (example, set job to run every Month, Mon to Friday at 11am sort of trigger)

Polling? Are you talking about the File Trigger? Then you should not use polling unless network server is a Samba share (Linux). In WIndows polling is not used and reaction is "instant".

Hi,
I just upload screen shot for the event trigger "polling" setting.



Definitely the write file tasks and read file and even the list file tasks we use (which seem to fail) correspond to a share (our windows boxes see them as shares, drive letter or a UNC path). The share physically correspond to a NAS (network attached storage server). It NAS (controller is a Linux box), presenting a CIF share.

What does the polling setting do?

I'm sorry. I think I'm making the problem description confusing.
Problem
- Jobs with tasks (Write file, List file, Read file), fail seemingly at random
- These jobs reside on multiple machines. Some machines have close to 250 VisualCron jobs

As an attempt to alert us of a failed task, we configured a job to check the outcome of the failed job. Example, if the job completed but had a failed task, send an email message and then write a file. We used the event trigger, VisualCron (see the previous screen attachements)

Likewise we also have a job which checks for jobs which have been inactivated, activated or deleted.

Think of these two jobs as "Checking type jobs". These two jobs are not failing per se. But we see a pattern where we believe these checking jobs are making the other jobs fail.

And the number occurrences of jobs with a failed task seem to increase as we decrease the polling frequency setting. Originally the checker type jobs, event trigger polling setting was set to 60 and 900 seconds. We increased the polling to 1800 seconds for both checker type jobs, and the environment has been running clean (no job failures) for about 24 hours.

We see a correlation between this polling setting and jobs (task failing).

Thanks for the information, we'll take a closer look at this. Do you see any correlation between the trigger you are using for the Write File, List File, Read File job/tasks? What trigger is being used for them?

Hi,
No I don't see a correlation with that trigger. The jobs with the failing tasks (write file, list file, read file) are triggered by the very common, CUSTOM TIME TRIGGER (example, set job to run every Month, Mon to Friday at 11am sort of trigger)

Thanks for the information

Are the tasks still failing or has they been running well since your polling update?

---

Polling? Are you talking about the File Trigger? Then you should not use polling unless network server is a Samba share (Linux). In WIndows polling is not used and reaction is "instant".

Hi,
I just upload screen shot for the event trigger "polling" setting.



Definitely the write file tasks and read file and even the list file tasks we use (which seem to fail) correspond to a share (our windows boxes see them as shares, drive letter or a UNC path). The share physically correspond to a NAS (network attached storage server). It NAS (controller is a Linux box), presenting a CIF share.

What does the polling setting do?

I'm sorry. I think I'm making the problem description confusing.
Problem
- Jobs with tasks (Write file, List file, Read file), fail seemingly at random
- These jobs reside on multiple machines. Some machines have close to 250 VisualCron jobs

As an attempt to alert us of a failed task, we configured a job to check the outcome of the failed job. Example, if the job completed but had a failed task, send an email message and then write a file. We used the event trigger, VisualCron (see the previous screen attachements)

Likewise we also have a job which checks for jobs which have been inactivated, activated or deleted.

Think of these two jobs as "Checking type jobs". These two jobs are not failing per se. But we see a pattern where we believe these checking jobs are making the other jobs fail.

And the number occurrences of jobs with a failed task seem to increase as we decrease the polling frequency setting. Originally the checker type jobs, event trigger polling setting was set to 60 and 900 seconds. We increased the polling to 1800 seconds for both checker type jobs, and the environment has been running clean (no job failures) for about 24 hours.

We see a correlation between this polling setting and jobs (task failing).

Thanks for the information, we'll take a closer look at this. Do you see any correlation between the trigger you are using for the Write File, List File, Read File job/tasks? What trigger is being used for them?

Hi,
No I don't see a correlation with that trigger. The jobs with the failing tasks (write file, list file, read file) are triggered by the very common, CUSTOM TIME TRIGGER (example, set job to run every Month, Mon to Friday at 11am sort of trigger)

Thanks for the information

Are the tasks still failing or has they been running well since your polling update?

Hi,
We have been running our environment with the adjusted polling setting for about 4 days. We have had three Visualcron jobs whose tasks have failed. The job failures have occurred on a machine which has the highest number of scheduled jobs. This is quite significant, because before we adjusted the polling setting, this machine would average 30 job failures each day. The other machines continue to run clean with no failures. Polling was also adjusted on these machines too.

I see. we looked at the code. When not using polling the polling interval is the interval that is used to update the folder and file mask - in case you use a Variable in any of these fields. If you have not checked polling and not use Variables you can use a very high value here.

---

I see. we looked at the code. When not using polling the polling interval is the interval that is used to update the folder and file mask - in case you use a Variable in any of these fields. If you have not checked polling and not use Variables you can use a very high value here.

Hi,
I'm sorry I'm not understanding your explanation.

The "Polling interval" is not an option (ie its not a check box), its a setting. Example: Check every xx seconds



Are you saying, because I use a variables for the a directory name or the name of the file.(example...\\servername\variable\variableFilename ..there is a conflict with the event trigger? Is this considered a bug? Or is it considered, working as designed?
If its considered working as designed, I would definitely like to see support for variables.

I see. we looked at the code. When not using polling the polling interval is the interval that is used to update the folder and file mask - in case you use a Variable in any of these fields. If you have not checked polling and not use Variables you can use a very high value here.

Hi,
I'm sorry I'm not understanding your explanation.

The "Polling interval" is not an option (ie its not a check box), its a setting. Example: Check every xx seconds



Are you saying, because I use a variables for the a directory name or the name of the file.(example...\\servername\variable\variableFilename ..there is a conflict with the event trigger? Is this considered a bug? Or is it considered, working as designed?
If its considered working as designed, I would definitely like to see support for variables.

I am talking about the Use polling checkbox.

You can use Variables in file mask. When Use polling is not checked the value is used to set how often to refresh the Variable. Think of it like this, you tell Windows to check for a value, a file name. Then you tell it to start. As you want to use a Variable that value can change any time, depending on what kind of Variable you are using. So, in general you do not want to tell Windows every second that now I want to look for another file, it will simply not be efficient. So, that is why we update the actual resolving of the Variable from time to time. Currently based on that value but it might change because it is not good that it is used for duplicate functions.

---