Would Multiple Processes Cause: luatex: execution interrupted
I run context in a task queue with four worker processes. It is possible that context is invoked simultaneously to process distinct input files by tasks running in separate worker processes. I see periodic fatal errors, as follows: mtx-context | fatal error: no return code, message: luatex: execution interrupted However, the same input files can be processed by context directly without errors. Is it possible that context running in simultaneous but separate processes conflict? This is LuaTeX, Version beta-0.70.1-2011051923 (rev 4277) ConTeXt ver: 2011.05.18 18:04 MKIV fmt: 2011.7.2 int: english/english
On 4 jul. 2011, at 23:24, Raymond LeClair
I run context in a task queue with four worker processes. It is possible that context is invoked simultaneously to process distinct input files by tasks running in separate worker processes. I see periodic fatal errors, as follows:
mtx-context | fatal error: no return code, message: luatex: execution interrupted
That is a 'hard' luatex crash. Because context calls luatex in a subprocess, it can be problematic to find the actual cause, unfortunately.
However, the same input files can be processed by context directly without errors. Is it possible that context running in simultaneous but separate processes conflict?
Quite possibly (but that is very hard to verify and would be even harder to debug). Best practise is to give each process its own temporary directory, just in case. Best wishes, Taco
This is LuaTeX, Version beta-0.70.1-2011051923 (rev 4277) ConTeXt ver: 2011.05.18 18:04 MKIV fmt: 2011.7.2 int: english/english
___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context webpage : http://www.pragma-ade.nl / http://tex.aanhet.net archive : http://foundry.supelec.fr/projects/contextrev/ wiki : http://contextgarden.net ___________________________________________________________________________________
Thanks for your reply! On Jul 4, 2011, at 5:35 PM, Taco Hoekwater wrote:
Quite possibly (but that is very hard to verify and would be even harder to debug). Best practise is to give each process its own temporary directory, just in case.
This is, in fact, guaranteed by design: each file resides in a directly which is named using a uuid. Is it possible that context or luatex are creating files in a common directory, say the cache directory, and access by one process locks the file for access by another?
Thanks for your reply!
On Jul 4, 2011, at 5:35 PM, Taco Hoekwater wrote:
Quite possibly (but that is very hard to verify and would be even harder to debug). Best practise is to give each process its own temporary directory, just in case.
This is, in fact, guaranteed by design: each file resides in a directly which is named using a uuid.
Is it possible that context or luatex are creating files in a common directory, say the cache directory, and access by one process locks the file for access by another? It's possible: say for example that each task rebuild the format at every run just to be sure. The format is shared by all instances for
On Tue, Jul 5, 2011 at 4:23 AM, Raymond LeClair
On 07/05/2011 08:11 AM, luigi scarso wrote:
Thanks for your reply!
On Jul 4, 2011, at 5:35 PM, Taco Hoekwater wrote:
Quite possibly (but that is very hard to verify and would be even harder to debug). Best practise is to give each process its own temporary directory, just in case.
This is, in fact, guaranteed by design: each file resides in a directly which is named using a uuid.
Is it possible that context or luatex are creating files in a common directory, say the cache directory, and access by one process locks the file for access by another? It's possible: say for example that each task rebuild the format at every run just to be sure. The format is shared by all instances for
On Tue, Jul 5, 2011 at 4:23 AM, Raymond LeClair
wrote: the same minimals or TeXLive, so conflicts may arise. The same for example for the font cache.
Possible but, I think, quite unlikely. Another possibility is that you simply run out of memory? Best wishes, Taco
Taco, Thanks for your reply! You are correct, the issue appears to be memory. The system is hosted on a linode, the smallest of which is 512 MB RAM. I increased this to 1024 MB RAM and this appears to fix the problem. Very much appreciate your help. Ray. On Jul 5, 2011, at 2:15 AM, Taco Hoekwater wrote:
On 07/05/2011 08:11 AM, luigi scarso wrote:
Thanks for your reply!
On Jul 4, 2011, at 5:35 PM, Taco Hoekwater wrote:
Quite possibly (but that is very hard to verify and would be even harder to debug). Best practise is to give each process its own temporary directory, just in case.
This is, in fact, guaranteed by design: each file resides in a directly which is named using a uuid.
Is it possible that context or luatex are creating files in a common directory, say the cache directory, and access by one process locks the file for access by another? It's possible: say for example that each task rebuild the format at every run just to be sure. The format is shared by all instances for
On Tue, Jul 5, 2011 at 4:23 AM, Raymond LeClair
wrote: the same minimals or TeXLive, so conflicts may arise. The same for example for the font cache. Possible but, I think, quite unlikely. Another possibility is that you simply run out of memory?
Best wishes, Taco
___________________________________________________________________________________ If your question is of interest to others as well, please add an entry to the Wiki!
maillist : ntg-context@ntg.nl / http://www.ntg.nl/mailman/listinfo/ntg-context webpage : http://www.pragma-ade.nl / http://tex.aanhet.net archive : http://foundry.supelec.fr/projects/contextrev/ wiki : http://contextgarden.net ___________________________________________________________________________________
participants (3)
-
luigi scarso
-
Raymond LeClair
-
Taco Hoekwater