Hi, is it possible that because of one of these properties I will experience the issues? ``` grpc.client.enable-keep-alive=true grpc.client.keep-alive-without-calls=true grpc.client.negotiation-type=plaintext ```
On Thursday, December 21, 2023 at 3:46:13 PM UTC+2 Noam Isachar wrote: > Thanks for the reply. My chunking helps because the large message contains > a lot of timeseries for forecasting, and if they exceed they maximum size I > can split them and send them in several messages instead of in a single > one. > Increasing the max size is not possible because my service crashes with > larger messages, so the only way here is to send messages lower than the > current maximum size. The problem is that when I check the message size > before sending it I get a size that should be good, but then I get a > Status.RESOURCE_EXHAUSTED with a different size than the one I checked with > before. > > On Wednesday, December 20, 2023 at 9:02:14 PM UTC+2 Larry Safran wrote: > >> Hi Noam, >> If you set CallOptions.withMaxOutboundMessageSize, then your write will >> get a Status.RESOURCE_EXHAUSTED based upon the actual size being sent which >> will be present in the description. You could catch that and then break >> your message apart when needed. >> >> I'm not clear how your chunking is helping compared to the chunking that >> is done at the network level. The example size you gave is much smaller >> than the 4MB default; have you tried using a larger max size? >> >> Thanks, >> Larry >> >> On Wed, Dec 20, 2023 at 8:32 AM Noam Isachar <noam.i...@gmail.com> wrote: >> >>> I'm using gRPC to send messages from Java to a Python service. Recently >>> I started getting larger messages that sometimes exceed the maximum message >>> size for gRPC. Increasing that size is not possible for me due to lack of >>> resources and timeouts, so I had to implement a chunking mechanism in order >>> to split the large messages. >>> >>> In Java, I check the message size using getSerializedSize() and if it >>> exceeds the maximum size, I do the chunking. However, I noticed that the >>> value I'm getting from getSerializedSize() is different and much smaller >>> than the one shown in the error "gRPC message exceeds maximum size" which >>> means that I still send messages that exhaust the resources. >>> >>> For example, how is it possible that getSerializedSize() returns 75274 >>> and the error shows a size of 179651? Both are supposed to show the size in >>> bytes. I also tried getting the size using the length of the byte array, >>> but it is identical to getSerializedSize(). >>> >>> How can I know the actual size gRPC will get before sending the message >>> so I can apply the chunking when I should? >>> >>> -- >>> You received this message because you are subscribed to the Google >>> Groups "grpc.io" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to grpc-io+u...@googlegroups.com. >>> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/grpc-io/602108bd-15fd-421a-96ad-4f7c52bf4a2en%40googlegroups.com >>> >>> <https://groups.google.com/d/msgid/grpc-io/602108bd-15fd-421a-96ad-4f7c52bf4a2en%40googlegroups.com?utm_medium=email&utm_source=footer> >>> . >>> >> -- You received this message because you are subscribed to the Google Groups "grpc.io" group. To unsubscribe from this group and stop receiving emails from it, send an email to grpc-io+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/grpc-io/be7504e3-32e0-44d6-b026-bfede8c811b1n%40googlegroups.com.
