Re: UNIX domain sockets MFC's

2007-05-15 Thread Oliver Fromme
Marc G. Fournier wrote: That's the fun part ... I can't seem to re-create it anywhere except that one server :( And it doesn't seem to matter how many jail(s) I have on it ... I just dump'd 25 jails off of it and onto another server, and its still rising ... Then maybe it's not

Re: UNIX domain sockets MFC's

2007-05-15 Thread Marc G. Fournier
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 - --On Tuesday, May 15, 2007 19:09:22 +0200 Oliver Fromme [EMAIL PROTECTED] wrote: If there isn't, then start the jails one after another (not all at once) and keep checking. Maybe it's just one specific jails (or a few of them) which trigger

Re: UNIX domain sockets MFC's

2007-05-14 Thread Robert Watson
On Sat, 12 May 2007, Marc G. Fournier wrote: The fix for this has now been merged as 1.155.2.22. As there have been no new reports of UNIX domain socket problems in the last couple of days, it sounds like the MFC of the last batch of fixes and cleanups has not lead to problems. I've just

Re: UNIX domain sockets MFC's

2007-05-14 Thread Marc G. Fournier
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 - --On Monday, May 14, 2007 11:29:12 +0100 Robert Watson [EMAIL PROTECTED] wrote: On Sat, 12 May 2007, Marc G. Fournier wrote: The fix for this has now been merged as 1.155.2.22. As there have been no new reports of UNIX domain socket

Re: UNIX domain sockets MFC's

2007-05-14 Thread Oliver Fromme
Marc G. Fournier wrote: I'm curious about something ... way back, when I was using unionfs, I had a major problem with vnode leakage ... as I mentioned before, this server is the only one I have that uses geom/gmirror on its drives, the rest all use hardware RAID ... is there

Re: UNIX domain sockets MFC's

2007-05-14 Thread Marc G. Fournier
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 - --On Monday, May 14, 2007 16:03:13 +0200 Oliver Fromme [EMAIL PROTECTED] wrote: FWIW, I have two servers running RELENG_6 (2 months old) using gmirror and with a few jails (not many, though ... they're used for Apache web servers and

Re: UNIX domain sockets MFC's

2007-05-12 Thread Robert Watson
On Fri, 11 May 2007, Abdullah Ibn Hamad Al-Marri wrote: The speed improvements associated with these MFC's is minor; primarily they are stability improvements under high load. There are major performance improvements in the 7.x implementation, especially for multi-core systems, but I have

Re: UNIX domain sockets MFC's

2007-05-12 Thread Abdullah Ibn Hamad Al-Marri
On 5/12/07, Robert Watson [EMAIL PROTECTED] wrote: On Fri, 11 May 2007, Abdullah Ibn Hamad Al-Marri wrote: The speed improvements associated with these MFC's is minor; primarily they are stability improvements under high load. There are major performance improvements in the 7.x

Re: UNIX domain sockets MFC's

2007-05-12 Thread Marc G. Fournier
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 - --On Friday, May 11, 2007 12:49:32 +0100 Robert Watson [EMAIL PROTECTED] wrote: On Tue, 8 May 2007, Robert Watson wrote: Right now I am tracking two known issues with UNIX domain sockets in RELENG_6: - Reported NULL point derference in

Re: UNIX domain sockets MFC's

2007-05-11 Thread Robert Watson
On Tue, 8 May 2007, Robert Watson wrote: Right now I am tracking two known issues with UNIX domain sockets in RELENG_6: - Reported NULL point derference in unp_connect(), which occurs due to the dropping of locks around sonewconn(). This is fixed in HEAD, and I am preparing an MFC of this

Re: UNIX domain sockets MFC's

2007-05-11 Thread Marc G. Fournier
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 - --On Friday, May 11, 2007 12:49:32 +0100 Robert Watson [EMAIL PROTECTED] wrote: On Tue, 8 May 2007, Robert Watson wrote: Right now I am tracking two known issues with UNIX domain sockets in RELENG_6: - Reported NULL point derference in

Re: UNIX domain sockets MFC's

2007-05-11 Thread Abdullah Ibn Hamad Al-Marri
On 5/11/07, Robert Watson [EMAIL PROTECTED] wrote: On Tue, 8 May 2007, Robert Watson wrote: Right now I am tracking two known issues with UNIX domain sockets in RELENG_6: - Reported NULL point derference in unp_connect(), which occurs due to the dropping of locks around sonewconn(). This

Re: UNIX domain sockets MFC's

2007-05-11 Thread Robert Watson
On Fri, 11 May 2007, Abdullah Ibn Hamad Al-Marri wrote: On 5/11/07, Robert Watson [EMAIL PROTECTED] wrote: On Tue, 8 May 2007, Robert Watson wrote: Right now I am tracking two known issues with UNIX domain sockets in RELENG_6: - Reported NULL point derference in unp_connect(), which

Re: UNIX domain sockets MFC's

2007-05-11 Thread Abdullah Ibn Hamad Al-Marri
On 5/11/07, Robert Watson [EMAIL PROTECTED] wrote: On Fri, 11 May 2007, Abdullah Ibn Hamad Al-Marri wrote: On 5/11/07, Robert Watson [EMAIL PROTECTED] wrote: On Tue, 8 May 2007, Robert Watson wrote: Right now I am tracking two known issues with UNIX domain sockets in RELENG_6: -

Re: UNIX domain sockets MFC's

2007-05-11 Thread Mark Linimon
On Fri, May 11, 2007 at 11:26:37PM +0300, Abdullah Ibn Hamad Al-Marri wrote: How safe is using 7.0 for MySQL server now? -CURRENT is never recommended for mission-critical applications. mcl ___ freebsd-stable@freebsd.org mailing list

UNIX domain sockets MFC's

2007-05-08 Thread Robert Watson
Dear all, I have just completed the merge of a significant number of UNIX domain socket bug fixes and cleanups from HEAD to RELENG_6. While my hope is that these are generally low-risk, and should improve UNIX domain socket stability and performance under load, there were quite a few