On Wed, May 18, 2022 at 09:40:03AM -0700, Ian Lance Taylor wrote: [...]
> Thanks for including the strace output. It shows that the clone > system call is failing with ENOSPC. There are several reasons that > clone can fail with ENOSPC. None of them have anything to do with > disk space (though shortage of inodes appears to be a possibility). > ESXI appears to map several internal errors to ENOSPC as well, mostly > dealing with lack of memory. > > You may need to discuss this with someone familiar with ESXI. Quick googling turns up [1] which supports what Ian said. It's interesting, whether there is a ESXi-native tool analogous to strace but which would provide "native" kernel error codes? That could possibly move us closer to the root cause of the problem. By the way, Harsh, in your original SO post you state that you | have the same program in python and C++, and they both run without any | issues in the same environment. Could you please show bits of this code? I'm asking because it's interesting whether you really do fork+exec (or something) like this in those programs. I would also be interesting to inspect the traces captured by strace of the (successful) runs of these programs to see whether the clone(2) syscall gets performed with the same arguments (I think the "flags" argument is of the most interest). Of course, this might be a red herring, and the problem might lie elsewhere, so the ability to somehow obtain the "native" kernel's error code would probably help a lot. 1. http://virtualization24x7.blogspot.com/2017/03/table-of-all-vmkernel-error-codes-in.html -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/20220519110101.gdbzf7rxf4m65bfu%40carbon.
