[Bug 249871] NFSv4 faulty directory listings under heavy load
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=249871 --- Comment #5 from Jason W. Bacon --- Disabling debug.vfscache didn't help. Adding the patch will take a little more doing, as this is a production system. I'll try reproducing the problem on a private test cluster first, maybe later this week. A little more info: The problem seems to be limited to the directory being hammered by multiple readers. Listing a sibling directory produces consistent and correct results. -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-bugs@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
Problem reports for b...@freebsd.org that need special attention
To view an individual PR, use: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=(Bug Id). The following is a listing of current problems submitted by FreeBSD users, which need special attention. These represent problem reports covering all versions including experimental development code and obsolete releases. Status |Bug Id | Description +---+--- New |197876 | [devfs] an error in devfs leads to data loss and New |198797 | [PATCH] Added an option to install BSDstats to bs New |202362 | ntp: restore refclocks selection (10.2-RELEASE re New |202740 | vi/ex string substitution problem when there is m New |204097 | witness_initialize() does not perform bound check New |206336 | [patch] usr.sbin/freebsd-update allow proxy confi New |209213 | UEFI Loader shows only black screen with Nvidia G New |210804 | installerconfig - using ZFS create in custom scri New |223470 | freebsd-update: Cannot identify running kernel (/ New |230620 | "install -d" issue New |235085 | [PATCH] Option to make rc.d/sysctl more verbose ( Open| 71667 | [patch] cleanup of the usr.sbin/bootparamd code Open|182466 | [headers] [patch] make self-contained Open|183618 | [panic] Dell PowerEdge R620 -- PERC H710 Mini (mf Open|187015 | agpgart: Panic make_dev_credv: bad si_name (error Open|192573 | Add ps(1) option: Print process start time in sec Open|194925 | [pf] [ifconfig] interface group keywords do not w Open|197921 | scheduler: Allow non-migratable threads to bind t Open|206528 | Emulex LPe 16002 FC HBA Not Recognized by oce(4) Open|206649 | cyapa(4): Add common gestures for Cypress APA I2C Open|207248 | [patch] daemon(8): Add option to redirect stdout Open|207940 | stand/efi/boot1: Add boot partition selection Open|212608 | sockstat(1) and lsof(8) can not identity the owne Open|220246 | syslogd does not send RFC3164-conformant messages Open|221305 | Mouse cursor loss when moving cursor while loadin Open|221550 | kern.bootfile returns only /kernel on mips64 (ERL Open|221854 | makefs: Reject UFS labels that are too long to fi Open|222632 | connect(2) not available in capability mode Open|226893 | freebsd-update: Support patchlevel argument for f Open|231810 | [build] release always fails with "mkimg: partiti Open|233578 | Unprivileged local user can prevent other users l Open|233988 | freebsd-update: Improve progress output on termin Open|236718 | system panics with message: vm_fault_hold: fault Open|237287 | moused(8) ignores button release events in virtua Open|237924 | Possible infinite loop in function empty_aux_buff Open|237981 | cxgb(4): Driver doesn't work with latest (7.12) C Open|238183 | cam/scsi/scsi_sa.c: warnings issued by static ana Open|238486 | Possible buffer overflow bug in sc_allocate_keybo Open|238550 | Touchpad (via SMBus) not working: Synaptics (SYN1 Open|238638 | mfi: Remove unnecessary pointer printing in mfi.c Open|238837 | init: Remove P_SYSTEM flag from PID 1 to allow ea Open|241697 | i915kms: Kernel panic loading module on custom ke Open|242212 | usr.sbin/mergemaster/mergemaster.sh: There is no Open|244181 | unzip: Add passphrase and GLIBC build support Open|244838 | "bectl activate -t" does not honor the -t flag in Open|247132 | Fix build error: use of undeclared identifier 'cp Open|247301 | linuxkpi: panic: Assertion td->td_lkpi_task == NU Open|248352 | mfi(4): Remove RAID map sync functionality Open|248715 | dhclient: prepend domain-name-servers in dhclient 49 problems total for which you should take action. ___ freebsd-bugs@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
[Bug 249871] NFSv4 faulty directory listings under heavy load
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=249871 --- Comment #6 from Rick Macklem --- One additional thing you could do is: # vmstat -z | fgrep mbuf_cluster on the server when it is under heavy load, to see if it is running out of mbuf clusters. (Theoretically the NFS server should keep working when mbuf clusters are exhausted, but the result would be a Readdir reply made up of a long list of regular mbufs. That could impact things like TSO, if the net interface on the server has that enabled.) Basically, other than possible mbuf exhaustion, I can't think of any way heavy load would affect the NFS server code (except slower response). Since the name cache doesn't seem to be the culprit, that leaves all the caching that goes on inside ZFS. --> If the readdir contents is somehow reordered by ZFS when the directory is under heavy readdir load or the directory offset cookies somehow change, that would explain the problem. Yet one more thing that you could try is having client mounts done with "nfsv3,rdirplus". You mentioned that NFSv3 worked ok. NFSv3 + ridrplus works more closely to NFSv4 in the server, and whether or not that fails might be useful information. -- You are receiving this mail because: You are the assignee for the bug. ___ freebsd-bugs@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
