amulyan13

This master is a week old. I will try the latest master. Below is the call that I am making. nfs_pwrite_async(nfscontexts[contextIdx % NUM_CONNECTIONS], hellofh, WriteData[bufIdx]->buffer, bytes_read /*size */, offset, write_callback,...

[writecorrupt.zip](https://github.com/sahlberg/libnfs/files/14572321/writecorrupt.zip) I have uploaded the traces of the corrupted write and its complete tcp stream. xid of the failing request is 0x9a27dbec. Thanks @sahlberg for looking at this.

Hi, I am unable to attach the packet traces, hence uploaded it to the drive and sharing the link here: https://drive.google.com/file/d/1wxZ9IVg-0LVGXE1_J5iH08ewG_oI6a11/view?usp=drive_link

Thanks @sahlberg Attaching my sample program here. [copylibnfs.txt](https://github.com/sahlberg/libnfs/files/14585277/copylibnfs.txt) >>Another question is, why are you using som many different contexts if you are single threaded and using the async api? >>...

Hi @sahlberg I tried commenting out rpc_write_to_socket(rpc) call as suggested but I still see the issue. >>* With a single threaded application and libnfs I doubt it will be beneficial...

Hi @sahlberg , Thanks for sharing the test program. With the above program I just have 1 connection going out to the server whereas I want multiple connections created to...

Tested this fix, this resolved the issue. Thanks.