NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
Hi, re: > After reviewing the configuration file, consolidation by revising the regular > expressions may not be a possibility because the product request statements > reference different upstream servers. If you are saying that all 66 REQUESTs are being made to different upstream LDMs, then this is the first time we have ever seen a configuration like this! re: > For this load, what would you recommend above and beyond the allocated 16GB > RAM > and single CPU (Xeon E-2640 v4)? There are a number of pieces of information that we would need before being able to give a good response to this question. Here are some questions and requests for output: 1) please send us the output of 'ldmadmin config' We want to see how large your LDM queue is, etc. 2) are you running metrics gathering? What we are referring to is running the following things from crontab: # # LDM-6.8.1+ metrics gathering # * * * * * bin/ldmadmin addmetrics > /dev/null 0 0 1 * * bin/ldmadmin newmetrics > /dev/null The metrics file created, ~ldm/var/logs/metrics.txt, will, after metrics gathering has run long enough, tell us important things as time series plots: age of the oldest product in the queue CPU loading etc. 3) we need to have some idea of what you are doing with the products that are being REQUESTed For instance, if they are being written to disk files, is the partition you are doing the writes on NFS mounted? If the file system to which received products are being written is slow, then the processing of the products will also be slow, and this _will_ affect how fast the LDM can be shutdown using 'ldmadmin stop'. Why? The LDM tries to let processes complete what they are doing before shutting down. If there are a number of processes waiting on system resources, it will take a long time for them to complete, and the sum of this kind of waiting adds up to the time it takes for the LDM to stop. Question: - is it not permissible to share more information about your setup so that we have the information needed to give good advice? For instance, my example about combining feed REQUESTs only focused on like products. REQUEST(s) for unlike products can also be combined if they will be coming from the same upstream LDM. - could you, for instance, share your setup with us in, for instance, a Google Meet? Comment: - we think that streamlining your LDM setup will go a LONG way towards mitigating the problems you are experiencing After reorganization (if possible, of course), an assessment of how much more computing resources are needed to accomplish the task(s) at had can be better made. Cheers, Tom -- **************************************************************************** Unidata User Support UCAR Unidata Program (303) 497-8642 P.O. Box 3000 address@hidden Boulder, CO 80307 ---------------------------------------------------------------------------- Unidata HomePage http://www.unidata.ucar.edu **************************************************************************** Ticket Details =================== Ticket ID: MDC-227212 Department: Support LDM Priority: Normal Status: Closed =================== NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.