This is the mail archive of the cygwin mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Cygwin services using uid 400, not SYSTEM. Why?


AARRGGHH! I have solved the problem... bloody tunnel vision :-(

The following will explain it in technocolor:

shaddy@bebop ~
$ ls -l /etc/inetd.conf
-rwxr-x--- 1 shaddy None 2434 Jul 10 04:52 /etc/inetd.conf

shaddy@bebop ~
$ chmod a+r /etc/inetd.conf

shaddy@bebop ~
$ ls -l /etc/inetd.conf
-rwxr-xr-- 1 shaddy None 2434 Jul 10 04:52 /etc/inetd.conf

shaddy@bebop ~
$ cygrunsrv -S inetd

shaddy@bebop ~
$ telnet localhost
Connected to bebop.
Escape character is '^]'.

CYGWIN_NT-5.1 1.5.20(0.156/4/2) (bebop) (tty1)


I'm so sorry I didn't pick up on this earlier. Thanks for your attention. If you have any ideas on the UID 400 problem, I'd still be very interested to hear what was happening on that.

On 7/19/2006 11:44 PM, Shaddy Baddah wrote:
Hi again,

On 7/19/2006 11:39 PM, Shaddy Baddah wrote:

This has happened to me a few times now :-( I don't know why, but the original email on the line just above, did not make it to the list (or the archive). For completeness, I've in-line included it below.


-------- Original Message --------
Subject: Re: Cygwin services using uid 400, not SYSTEM. Why?
Date: Wed, 19 Jul 2006 23:39:00 +1000
From: Shaddy Baddah <>
References: <> <>


On 6/2/2006 2:35 AM, Shaddy Baddah wrote:
> Here is a break-down of the bizarre behaviour I am experiencing:
> 1. After installing the service, and starting it, the inetd (and its'
> wrapping cygrunsrv) processes was listed with ps as having UID 400.
> I'm at my wits end. Any advice would be greatly appreciated.

Sorry for the trickle of emails on this subject, but I am still facing
problems, and have been a little inundated with work recently.

I have an update on this. The problem I described was not only a problem
for inetd, but for ssh and exim as well.

Well, after running cygserver-config, and before running the cygserver
service itself, I was able to run the ssh and exim services (as before),
and actually make valid connections to the servers (as was not happening
before). I was also able to run the cygserver as well, and it did not
interfere negatively with the aforementioned services.

However, inetd service remains recalcitrant. I can run the service, but
I can make absolutely no connections to any of the served ports. I can
see the process in the process list.

In a vain attempt to try and diagnose the problem, I tried to attach
with strace -p, but that caused a deadlock situation.

Running inetd outside of a cygrunsrv service works ok as described in my
previous emails. I'm limping along in this way.

I'm not complaining, because if I were on the other side of this
correspondence, I could see that this might involve me a drawn out email
exchange, and be hesitant to reply, as appears to have been the case.
However, I am willing to try and work out the problem myself.

I just need some guidance please. When it comes to working with service
processes, things seem to be quite different. I can try to strace
process to some effect (indeed, I wrapped the command-line in an
strace... however, I am much better versed in truss under Solaris), but
the whole service framework throws my technique.

I've contemplated compiling inetd to put loads of logging, so that I can
see where it's getting to. But that seems drastic. I'm willing to take
the measure, but I hope this email my elicit a helpful preventative

Thanks in advance,

-- Unsubscribe info: Problem reports: Documentation: FAQ:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]