span8
span4
大家好,我工作的一个FME工作台为将他们的FME Server上运行的客户端。
一些背景资料:
我得到两个主要问题(当它不是一个,它的除外):
第一和因果错误(从我所知)是这一个:我得到的是旧“Insufficient memory available -- error code was 2 - please read the FME Help section 'Adjusting Memory Resources' for workarounds.“我已经与修修补补的FME_MEMORY_REDLINE无果的设置。
一旦在屏幕会闪烁,过程结束,不过,我无法利用日志。FME服务器好像删除它们没有备份,一旦任务失败,所以我无法挖掘并了解更多的细节。他们应该是在C:/ ProgramData /安全软件/ FME服亚搏在线务器/资源/日志/发动机/
这里是真正奇怪的事情:我看了运行中间夹上一次,并注意到两个陌生的行为。
在FME Server的网络信息中心,开始时间会每隔15分钟左右的变化,以及日志会显示什么似乎是工作台重新启动而不发出一个错误。唯一的地方,这是拍摄的是在状态窗格,因为日志远去,这并不表明在第15分钟,开始时间显示出一样的提交时间,然后更新18:45左右,然后 updated again at 19:00:
当这种变化发生时,我上了FME服务器框中,检查日志文件夹,发现这个(预期):
然后job_93文件夹内,有这样的:
有一个子文件夹,并在运行过程中另一个日志文件,只需在日志中开始时间的变化和明显的重新启动后,但在作业结束(原因不明,可能是内存问题),它抹去日志而不备份到/旧的文件夹。
我不能确定这多少明显的重启/日志抹行为是某种形式的预期行为,但它使调试相当困难。
至于内存不多了,没有任何人有除调整FME_MEMORY_REDLINE建议?而且,随着FME_TEMP文件夹这么大的空间,为什么不能在本地缓存和直接跳跃到内存中运行了呢?
更新:
与呼叫后@hollyatsafe从安全的其他一些人,我们似乎它的工作!亚搏在线日志失踪的问题似乎是侥幸,既然还没有发生。内存问题可能仍然存在问题,但我们已经想出了办法通过运行特征读者数据集(而不是普通的老读者)的空间裁剪小节来缓和。
如果这些问题持续存在,我会复活这个线程,但现在我们似乎已经在控制之下。感谢那些谁张贴的答案,并以安全乡亲直接帮助!亚搏在线
工作基于“增大化现实”技术e only resubmitted on FME Server if the translation ends unexpectedly, when the connection to an Engine is lost.This is most commonly due to an engine crash, therefore checking the fmeprocessmonitorengine.log as Jen suggested would confirm this, looking for errors similar to 'process engine_1 ended unexpectedly.'
In these circumstances the default resubmit value is 3, so this is why you are seeing a change in the timestamps.There should be a log created for each job run, located in the job id folder.I suspect the Web UI is looking for the final job log, and since there is only a log for the first job (_0) in this folder this is why Web UI says the job log does not exist, however since the timestamps are evenly 15 minutes apart this would suggest the memory is always peaking in the same place so reviewing the Job93_0.log should provide a good indication of where in the translation fmeengine is crashing.
As some of the logs do appear to be missing I'd recommend checking the FME Server clean up tasks to make sure these aren't being removed.If you'd prefer you can also change the FME Server configuration to stop resubmits altogether by editing the MAX_FAILED_TRANSACTION_REQUEST_RETRIES in thefmeserverconfig.txt(http://docs.safe.com/fme/html/FME_Server_Documentation/AdminGuide/Job_Recovery.htm).
Lastly if this is due to an engine crash you could also take a look in Windows Event Viewer over this timestamp which might provide some additional information.
As for the memory error, I actually encountered these exact same symptoms last week, unfortunately I was only able to resolve the issue by increasing the RAM.
嗨@ttaylor43
What version of FME Server are you using?In more recent versions of FME, altering the redline does next to nothing for controlling memory usage.
If the job fails (like it is for you) the engine will re-submit it, in the folder for job_93 you should see three logs with the suffix _0 _1 _2 for each resubmit.Then the view job should show the final log.
One place you could check would be the fmeserver.log file and the processmonitorengine.log - do those have any clues in there?
©2020安全亚搏在线软件公司|法律