This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Stacey, > We are having and LDM issue that I'm hoping you might be able to help us > with. We are placing data on the LDM queue for a machine and that > machine writes the data out to files. What we are seeing is a lag of > 8-10 minutes or more from when the data is received on the LDM queue to > when it actually gets written out to disk. Would you happen to know > what might be causing this lag? The machine does receive a lot of data > on the LDM queue, but it seems to work fine some times and then take > 8-10 minutes to write the files out at other times. We are stumped on > this one. Have you seen this behavior before? Do you know what might > be causing it? Any help you can provide would be greatly appreciated. Assuming that the data-products are being written to files using the FILE action of pqact(1), then the reason for the delay could very well be caused by the throughput through the LDM product-queue. If you're receiving a lot of data-products (e.g., all of CONDUIT) and doing a lot of processing of that data, then when the data arrives at a high rate (which CONDUIT does periodically) the system will be so busy that it might take some minutes to FILE a just-added data-product. For this reason, we recommend that LDMs that request a lot of data don't process that data locally but, instead, pass the data on to other, requesting LDMs that process subsets of the data. Hope this helps. > Thank you, > Mr. Stacey Simmons > WeahterTAP.com Regards, Steve Emmerson Ticket Details =================== Ticket ID: BLB-756234 Department: Support LDM Priority: Normal Status: Closed