Jobs not running after a while - VisualCron

Please note that VisualCron support is not actively monitoring this community forum. Please use our contact page for contacting the VisualCron support directly.

Welcome Guest! To enable all features please Login or Register.

zeleni
No customer Topic Starter

14 years ago

Hi

I use VC to run some php files in the bacground, updating different information in databases.

I have about 20 active jobs, that run with different frequencies, some rune once every day, some run every minute.

A job that is suposed to fetch some log files for parsing, just gives up after a while, it varies a lot. It can take an hour for it to do it, it can take a week, it is very random. There are no preconditions for running these jobs, only time aspect.

Any thoughts about what do i need to check ?

Sponsor

Forum information

Wanna join the discussion?! Login to your Forum forum account or Register a new forum account

Support
Official support

14 years ago

What are you saying;

1. That the next run property on the Job is "in the history"

or

2. That the next run is correct, but the Job do not start?

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

zeleni
No customer Topic Starter

14 years ago

The result is success on the last run, but the time in the next run field is way past.

Like few minutes ago, just before i made post on the forums, next run field was 08:54 on couple of jobs that run quite often, while all the other were normal, ie 10:30 something and were running properly.

Just happened again actually. I have the latest version of VC.

EDIT:

Just noticed in the VC log viewer, that i am getting an error every so often.

thisTimeTick was null in TimePicker

Error is not connected to any job.

Edited by user 14 years ago | Reason: Not specified

Support
Official support

14 years ago

1. So, it happens frequently and can be reproduce quite fast?

2. are you using more than one Trigger on each Job?

3. what kind of trigger - please post screenshots of Trigger settings.

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

Support
Official support

14 years ago

Ok, interesting. This might indicate some hardware problem.

Have you tested VisualCron with same settings on another machine?

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

zeleni
No customer Topic Starter

14 years ago

Support wrote:

1. So, it happens frequently and can be reproduce quite fast?

2. are you using more than one Trigger on each Job?

3. what kind of trigger - please post screenshots of Trigger settings.

1. Happened twice today, so i guess it can be reproduced in short time.

2. Only one trigger on the jobs.

3. Run every 30 seconds.

Support wrote:
Ok, interesting. This might indicate some hardware problem.

Have you tested VisualCron with same settings on another machine?

This is a vmware enviroment, the vc is running on a guest OS. I could replicate the virtual guest, but it would be running on the same hardware, so pretty pointless.

Any thoughts on what kind of hardware problem ?

Any thoughts on why would the next run parameter for a job stop updating ?

Edited by user 14 years ago | Reason: Not specified

Support
Official support

14 years ago

Quote:

This is a vmware enviroment, the vc is running on a guest OS. I could replicate the virtual guest, but it would be running on the same hardware, so pretty pointless.

I see, I thought you had another "real" server in your environment.

Quote:

Any thoughts on what kind of hardware problem ?

Not sure, could be something with VMWare as well. But this is a central part of VisualCron which queries the current time a lot.

Quote:

Any thoughts on why would the next run parameter for a job stop updating ?

What is happening is that we receive null values when querying for the time. This results in that seconds are "skipped". As your Job needs to be executed on a certain second (all do) it fails to make that match because the second was "skipped".

We'll see if we can create a test application to confirm this.

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

zeleni
No customer Topic Starter

14 years ago

Any chance to flag the time acquire failure, and then poll the time again in few seconds, instead of just giving up the next run update ?

The time is set through the netnod NNTP on all nodes in my network, as some functions are pretty dependant on the accuracy of time.

Edited by user 14 years ago | Reason: Not specified

Support
Official support

14 years ago

This is a central part which has not failed in years. We do not query NNTP servers but just the time of the computer.

Let us make a test and then maybe we can see the problem more clearly.

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

Support
Official support

14 years ago

#10

How often do you see this in your log file?

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

Support
Official support

14 years ago

#11

I am attaching an archive. Extract. Run and click on Start. Let it wait and we will see if any error arises. Please post any error here.

File Attachment(s):

Debug.zip (20kb) downloaded 49 time(s).

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

zeleni
No customer Topic Starter

14 years ago

#12

The errors occur anything from 30 sec to 2 min in frequency. Most likely when some of the jobs need to run.

I have downloaded the application and started it. The start button is grayed out, what is suposed to happen ?

I have temporarily worked around the issue by starting up the "run at random" function on the jobs that give up.

Edited by user 14 years ago | Reason: Not specified

zeleni
No customer Topic Starter

14 years ago

#13

I have been running your little application whole night, no errors. I supose they would be printed out below the start button ? Cant see any log files either.

Support
Official support

14 years ago

#14

A popup should show if any problems arise. No logs are created.

Is the VisualCron service having a high cpu usage?

Could you please turn on "Extended debugging" in Server settings then send the log_serverDATE.txt after the error has happened again.

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

zeleni
No customer Topic Starter

14 years ago

#15

No popups during the night.

One job stopped @ 3am today.
One @ 8.15am today.

Started extended debuging. Cant see any server log files with DATE in them, guess they havent been created yet.

Support
Official support

14 years ago

#16

c:/program files/VisualCron/log

log_serverDATE.txt

DATE is an example of the current date.

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

zeleni
No customer Topic Starter

14 years ago

#17

Here is the log file.

Job fetch_files had next run at 10:25 where it stopped and did not run again.

Also job import_mx1_alarmlog last run was suposed to be at 11:01, but it did not run again. I hope this helps.

Edited by user 14 years ago | Reason: Not specified

File Attachment(s):

log_server20101216.zip (216kb) downloaded 50 time(s).

Support
Official support

14 years ago

#18

Strange,

I don't see the error you mentioned in this log file. Could you export the settings (File->Export) and send to support@visualcron.com?

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

zeleni
No customer Topic Starter

14 years ago

#19

There are no errors, except the timepicker one.

Config sent via email.

Support
Official support

14 years ago

#20

Thanks for the files. I did not see the timepicker errors in your file log?

Could you try changing triggers to custom triggers instead? So that you use the following, for 30 seconds for example:

Every year, Every day, every month, every hour, every minute, second 0 and 30

It would be interesting if you still get that error then.

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

zeleni
No customer Topic Starter

14 years ago

#21

I have updated the 2 jobs that go inert most often with the custom time triggers.

Ill check upon it in few hours.

Kinda weird that you didnt see the picker errors.

2010-12-16 00:01:08 Err thisTimeTick was null in TimePicker

2010-12-16 04:24:08 Err thisTimeTick was null in TimePicker

2010-12-16 10:30:43 Err thisTimeTick was null in TimePicker

These are just 3 of may hundered in the log.

Edited by user 14 years ago | Reason: Not specified

Support
Official support

14 years ago

#22

Yes, we saw them later, we did search for the wrong text.

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!

zeleni
No customer Topic Starter

14 years ago

#23

After i switched to custom time triggers, seems to be working. I guess those are not dependent on time polling ? Bit more hassle to set up, but at least it works now.

Thank you for great help.

Support
Official support

14 years ago

#24

Thanks for the update - we will try to reproduce this.

Henrik
Support
http://www.visualcron.com
Please like VisualCron on facebook!