On OI (and Solaris 11.4), nscd caches auth_attr, user_attr, prof_attr, and 
exec_attr (and of course passwd, group, and project); so I'm not sure what 
could be done to speed up RBAC. However, policy.conf is NOT cached (and has 
lots of CDDL etc comments in it); per policy.conf(4) it's read by 
chkauthattr(3SECDB) and getexecuser(3SECDB).

OTOH, unless one is using one of the pf*sh shells or pfexec(1), I don't see how 
RBAC is part of the problem anyway.

> On Jan 19, 2021, at 05:17, Jim Klimov <[email protected]> wrote:
> 
> On January 19, 2021 4:27:04 AM UTC, Hung Nguyen Gia via openindiana-discuss 
> <[email protected]> wrote:
>> OI included many shells. So far I only stick with the default, bash.
>> 
>> But the performance is very bad. e.g: when using pkgsrc to build
>> packages.
>> 
>> As I have said on this list: pkgsrc bootstrap on OI is 4x slower than
>> on FreeBSD. Building packages also that slower.
>> 
>> I think the problem is of the shell. Because I see it checking for
>> something very slowly.
>> 
>> The output printed on the screen 'Checking for...' is line by line,
>> very slow. Meanwhile, the same thing on FreeBSD is blazing fast that I
>> can't even see what's going on at all.
>> 
>> I'm thinking about using other shell than bash.
>> 
>> But I can't test with each shell. They are too many.
>> 
>> From your own experience, which shell is the fastest?
>> 
>> _______________________________________________
>> openindiana-discuss mailing list
>> [email protected]
>> https://openindiana.org/mailman/listinfo/openindiana-discuss
> 
> This matches my experience sadly, also on systems with ksh93 as the real 
> default system shell since Solaris, and e.g. configure scripts using that 
> as-is or patched to use bash explicitly.
> 
> Same codebase mounted from OI over NFS to a Linux VM passes configuration 
> much faster - so it is not e.g. overheads of disk/FS layers.
> 
> Similarly for shell/fork heavy tests like https://github.com/42ity/JSON.sh 
> unit-testing (nearly zero I/O, but lots of shells tried) - the linux worker 
> completes minutes before OI does.
> 
> As far as I could unravel and guess, this is just about a more expensive 
> forking routine (RBAC and all) than on less protective OSes. This is a PITA 
> sadly, but unless something is just broken in the kernel but rather really 
> does more work by design because of different goals and trade-offs, then so 
> be it.
> 
> If something *is* broken and can be made faster, it would be much appreciated 
> :-)
> 
> Jim
> 
> --
> Typos courtesy of K-9 Mail on my Android
> 
> _______________________________________________
> openindiana-discuss mailing list
> [email protected]
> https://openindiana.org/mailman/listinfo/openindiana-discuss
> 

_______________________________________________
openindiana-discuss mailing list
[email protected]
https://openindiana.org/mailman/listinfo/openindiana-discuss

Reply via email to