What does -1 for start mean in a session?
I am seeing some scraping sessions just kinda hanging, leaving 0 length log files, and when I (using SOAP) view the list of sessions on that server the "start" is not a time (like 1176758577171) but a -1. I cannot (via my soap interface, again) stop or remove the sessions.
Any ideas?
What does -1 for start mean in a session?
Hi,
Is there any chance screen-scraper could be running out of memory? It can often exhibit unusual behavior when this occurs. Is there any chance that increasing the memory allocation to screen-scraper might help?
Thanks,
Todd
What does -1 for start mean in a session?
I like that explanation, and for the most part it seems to hold up, but I've got an item that's been just sitting there (sometimes all alone) in the running sessions list with no change for the bulk of the day now. And scrapes of the same name (with relatively similar parameters) have come and gone long ago.
Appreciate my anecdotal evidence? The lack of details? Or even the generalizations? I know how sketchy questin is, I was hoping for a "Oh yeah sometimes the migration bit seques into the index queue and it transposes the log file nomenclature format, just make sure you're adjusting the flange coupling before you call the interlocutor." Or some other kind of known issue.
I'll mess with the settings, but I know we're had tons of scrapes passing through at much higher volumes than these are at. Wrapper log has these complaints:
ERROR | wrapper | 2007/04/18 10:32:57 | JVM did not exit on request, terminated
ERROR | wrapper | 2007/04/18 11:21:42 | JVM appears hung: Timed out waiting for signal from JVM.
ERROR | wrapper | 2007/04/18 11:21:43 | JVM did not exit on request, terminated
But those are after this session would have begun.
I will also upgrade that server to the latest alpha version so I can use that nifty web interface.
What does -1 for start mean in a session?
Hi,
There's a chance those scraping sessions could be simply queued up to run. Depending on how many scraping sessions you've designated to run concurrently, it may be leaving some in the queue, but would run them eventually. You can control how many scraping sessions screen-scraper will run simultaneously in the "Settings" window (click the wrench icon) under the "Maximum number of concurrent running scraping sessions" setting.
Feel free to post back if that doesn't help.
Kind regards,
Todd Wilson