Failed to start daemon

How can i diagnose further the failed daemon startup.
I am using Centos 7 host and connecting to Centos 7 target via ssh. Nsight Systems comming with cuda toolkit 11.2 - 2020.4.3
DaemonStartError (1405) {
OriginalExceptionClass: N5boost16exception_detail10clone_implIN13QuadDAnalysis16DaemonStartErrorEEE
OriginalFile: /build/agent/work/20a3cfcd1c25021d/QuadD/Host/Analysis/SshDevice.cpp
OriginalLine: 1199
OriginalFunction: virtual std::__cxx11::string QuadDAnalysis::SshDevice::StartDaemon(const string&)
ErrorText: Daemon start failed.
ExitCode: 1

Hi Jacek, could you first try our latest release (2021.2)? It can be downloaded from Download Center | NVIDIA Developer.

What do you mean by “first”? Providing a log location seems quicker than installing new release? Does it require newer cuda toolkit as well? and drivers?

2020.4 has been several releases behind our latest update, so there may be known bugs that have already been fixed. We generally recommend trying latest release before digging further.

It does not require new CTK or drivers.

The installer does not work:
$ ./
Verifying archive integrity… 100% All good.
Uncompressing NVIDIA Nsight Systems 100%
Can’t locate in @INC (@INC contains: /usr/local/lib64/perl5 /usr/local/share/perl5 /usr/lib64/perl5/vendor_perl /usr/share/perl5/vendor_perl /usr/lib64/perl5 /usr/share/perl5 .) at ./ line 10.
BEGIN failed–compilation aborted at ./ line 10.

After doing

yum install perl-Env

the installation finishes ok. And 2021.2.1.58 does not have this issue (i.e. ssh target works fine). Thank you.

1 Like