Re: High rate of NFS cache misses after upgrading from 10.3-prerelease to 11.1-release
On 04/17/2018 01:03 AM, Rick Macklem wrote: > Niels Kobschaetzki wrote: > [stuff smipped] >> I just checked the code to see if I can figure out where exactly I have >> to put the printf(). And then I saw that there are ifdefs for >> NFS_ACDEBUG which seems to be a kernel option. When I add NFS_ACDEBUG in >> the config-file for the kernel, the build fails with an > I don't have sources handy right now, but you can probably just put a line > like: > #define NFS_ACDEBUG 1 > at the top of the file /usr/src/sys/fs/nfsclient/nfs_clsubs.c ok > After building/booting the kernel "sysctl -a" should have a > vfs.nfs.acdebug > in the list. Set it to "1" to get the basic timeout info. > >> /usr/src/sys/amd64/conf/ACDEBUG: unknown option "NFS_ACDEBUG" >> >> I looked in sysctl.h and there it isn't defined. Do I do something wrong >> or did this sysctl-tunable got lost at some point in time? >> Can I just use this code by removing the ifdef for getting information? >> >> Sorry, my C is not really existent, thus I have to ask :/ >> I would remove the ifdefs and the "if (nfs_acdebug …)" > This would work, too, rick That worked, I had the kernel running yesterday. I solved now finally my problem after two weeks and it wasn't the NFS. I just got derailed from the real solution again and again from some people, thus I didn't look in the right place. The cache misses are gone now, the application performs now faster than on the other servers. Thanks so so much for your help. Niels ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
Re: High rate of NFS cache misses after upgrading from 10.3-prerelease to 11.1-release
Niels Kobschaetzki wrote: [stuff smipped] >I just checked the code to see if I can figure out where exactly I have >to put the printf(). And then I saw that there are ifdefs for >NFS_ACDEBUG which seems to be a kernel option. When I add NFS_ACDEBUG in >the config-file for the kernel, the build fails with an I don't have sources handy right now, but you can probably just put a line like: #define NFS_ACDEBUG 1 at the top of the file /usr/src/sys/fs/nfsclient/nfs_clsubs.c After building/booting the kernel "sysctl -a" should have a vfs.nfs.acdebug in the list. Set it to "1" to get the basic timeout info. >/usr/src/sys/amd64/conf/ACDEBUG: unknown option "NFS_ACDEBUG" > >I looked in sysctl.h and there it isn't defined. Do I do something wrong >or did this sysctl-tunable got lost at some point in time? >Can I just use this code by removing the ifdef for getting information? > >Sorry, my C is not really existent, thus I have to ask :/ > >The parts (except the part that looks at the sysctl looks like this): >#ifdef NFS_ACDEBUG >if (nfs_acdebug>1) > printf("ncl_getattrcache: initial timeo = %d\n", timeo); >#endif > >…… > > >#ifdef NFS_ACDEBUG >if (nfs_acdebug > 2) >printf("acregmin %d; acregmax %d; acdirmin %d; acdirmax >%d\n", >nmp->nm_acregmin, nmp->nm_acregmax, >nmp->nm_acdirmin, nmp->nm_acdirmax); > >if (nfs_acdebug) >printf("ncl_getattrcache: age = %d; final timeo = %d\n", >(time_second - np->n_attrstamp), timeo); >#endif > > >I would remove the ifdefs and the "if (nfs_acdebug …)" This would work, too, rick ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
Re: QOS/TOS in carp/freebsd in general
On Mon, Apr 16, 2018 at 1:27 AM, Tom Joneswrote: > On Sun, Apr 15, 2018 at 04:30:45PM -0400, Nick Wolff wrote: > > Hi, > > > > I currently have a review in to make carp use dscp values on outgoing > > packets. This will make it easier to preform QOS on modern switches as we > > have been able to drive carp insane on 10g links while throwing storage > > traffic at it.On an interesting side note apparently after 64k mac > address > > moves for a single mac address in the cam table arista 7050t seems to > with > > out warning give up and stop moving address but leaves it in the table > > where ever it last was. > > > > I was originally going to just toggle between cs7 and the old TOS low > delay > > setting. But it was requested that I just make it a settable Integer > value. > > In this case I'm planning to do 0-63 to match dscp with a default of 54 > > (CS7) (Network Control) but you could still set the old value by setting > > the sysctl to 4. Anything larger then 64 would be truncated as two shifts > > are needed to align with the field leading. > > > > Does anybody do anything anywhere with the old TOS values like low delay > in > > there networking infrastructure? Should this be updated in other places > in > > kernel and userland? It's been 20 years since TOS was deprecated and > > replaced with DSCP. > > > > Any other comments or thought are always welcome. > > You can already do this with ipfw? setdscp cs7 from me to 224.0.0.18 ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
[Bug 227461] [hast] regression, unable to start hast, pid XXX (hastd), uid 0: exited on signal 6 (core dumped)
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=227461 Eugene Grosbeinchanged: What|Removed |Added CC||eu...@freebsd.org, ||n...@freebsd.org Assignee|b...@freebsd.org|p...@freebsd.org -- You are receiving this mail because: You are on the CC list for the bug. ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
Re: High rate of NFS cache misses after upgrading from 10.3-prerelease to 11.1-release
On 04/14/2018 03:49 AM, Rick Macklem wrote: > If you go into ncl_getattrcache() {it's in sys/fs/nfsclient/nfs_clsubs.c} > and add a printf() for "time_second" and "np->n_mtime.tv_sec" near the > top, where it calculates "timeo" from it. > Running this hacked kernel might show you if either of these fields is bogus. > (You could then printf() "timeo" and "np->n_attrtimeo" just before the "if" > clause that increments "attrcache_misses", which is where the cache misses > happen to see why it is missing the cache.) > If you could do this for the 10.3 kernel as well, this might indicate why the > miss rate has increased? I just checked the code to see if I can figure out where exactly I have to put the printf(). And then I saw that there are ifdefs for NFS_ACDEBUG which seems to be a kernel option. When I add NFS_ACDEBUG in the config-file for the kernel, the build fails with an /usr/src/sys/amd64/conf/ACDEBUG: unknown option "NFS_ACDEBUG" I looked in sysctl.h and there it isn't defined. Do I do something wrong or did this sysctl-tunable got lost at some point in time? Can I just use this code by removing the ifdef for getting information? Sorry, my C is not really existent, thus I have to ask :/ The parts (except the part that looks at the sysctl looks like this): #ifdef NFS_ACDEBUG if (nfs_acdebug>1) printf("ncl_getattrcache: initial timeo = %d\n", timeo); #endif …… #ifdef NFS_ACDEBUG if (nfs_acdebug > 2) printf("acregmin %d; acregmax %d; acdirmin %d; acdirmax %d\n", nmp->nm_acregmin, nmp->nm_acregmax, nmp->nm_acdirmin, nmp->nm_acdirmax); if (nfs_acdebug) printf("ncl_getattrcache: age = %d; final timeo = %d\n", (time_second - np->n_attrstamp), timeo); #endif I would remove the ifdefs and the "if (nfs_acdebug …)" Niels ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
[Bug 221317] Netmap issue after ixgbe driver update in r320897
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=221317 --- Comment #29 from Sylvain Galliano--- (In reply to Stephen Hurd from comment #28) Patch with error logs applied: I do not have any error log before issue to appear. -- You are receiving this mail because: You are on the CC list for the bug. ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
[Bug 194515] [pf] [vnet] Fatal Trap 12 Kernel with vimage
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=194515 Kubilay Kocakchanged: What|Removed |Added Keywords|needs-qa|crash -- You are receiving this mail because: You are on the CC list for the bug. ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
Re: QOS/TOS in carp/freebsd in general
On Sun, Apr 15, 2018 at 04:30:45PM -0400, Nick Wolff wrote: > Hi, > > I currently have a review in to make carp use dscp values on outgoing > packets. This will make it easier to preform QOS on modern switches as we > have been able to drive carp insane on 10g links while throwing storage > traffic at it.On an interesting side note apparently after 64k mac address > moves for a single mac address in the cam table arista 7050t seems to with > out warning give up and stop moving address but leaves it in the table > where ever it last was. > > I was originally going to just toggle between cs7 and the old TOS low delay > setting. But it was requested that I just make it a settable Integer value. > In this case I'm planning to do 0-63 to match dscp with a default of 54 > (CS7) (Network Control) but you could still set the old value by setting > the sysctl to 4. Anything larger then 64 would be truncated as two shifts > are needed to align with the field leading. > > Does anybody do anything anywhere with the old TOS values like low delay in > there networking infrastructure? Should this be updated in other places in > kernel and userland? It's been 20 years since TOS was deprecated and > replaced with DSCP. > > Any other comments or thought are always welcome. I cannot speak to network stacks, but we have performed some large scale DSCP measurements of the Internet and found TOS semantics still in use. You might want to look up: Exploring DSCP modification pathologies in mobile edge networks: IEEE/IFIP Workshop on Mobile Network Measurement I think there is more work in line to be published based on similar measurement campaigns. > References: > > Open Review - https://reviews.freebsd.org/D14536 > General DSCP reference - https://www.tucny.com/Home/dscp-tos > > > Thanks, > > Nick "darkfiberiru" Wolff > ___ > freebsd-net@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-net > To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org" -- - [tj] ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
[Bug 194515] [pf] [vnet] Fatal Trap 12 Kernel with vimage
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=194515 --- Comment #23 from Eugene Grosbein--- It seems the problem was fixed with r302156 before stable/11 was branched but never merged to stable/10. -- You are receiving this mail because: You are on the CC list for the bug. ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
[Bug 194515] [pf] [vnet] Fatal Trap 12 Kernel with vimage
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=194515 Eugene Grosbeinchanged: What|Removed |Added Assignee|n...@freebsd.org |b...@freebsd.org CC||n...@freebsd.org -- You are receiving this mail because: You are the assignee for the bug. You are on the CC list for the bug. ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
[Bug 194515] [pf] [vnet] Fatal Trap 12 Kernel with vimage
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=194515 Eugene Grosbeinchanged: What|Removed |Added See Also|https://reviews.freebsd.org |https://reviews.freebsd.org |/D1944 |/D6924 -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
[Bug 194515] [pf] [vnet] Fatal Trap 12 Kernel with vimage
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=194515 Eugene Grosbeinchanged: What|Removed |Added CC||eu...@freebsd.org Summary|Fatal Trap 12 Kernel with |[pf] [vnet] Fatal Trap 12 |vimage |Kernel with vimage See Also||https://reviews.freebsd.org ||/D1944 -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"