Re: ZFS secondarycache on SSD problem on r255173
On Fri, Apr 25, 2014 at 08:26:52PM -0500 I heard the voice of Matthew D. Fuller, and lo! it spake thus: On Mon, Mar 03, 2014 at 11:17:05AM +0200 I heard the voice of Andriy Gapon, and lo! it spake thus: I noticed that on some of our systems we were getting a clearly abnormal number of l2arc checksum errors accounted in l2_cksum_bad. [...] I propose the following patch which has been tested and seems to fix the problem without introducing any side effects: I've been running this for 2 weeks now without any cksum_bad's showing up (long enough that without it, I'd have expected them), and no other obvious side effects. Another 6 weeks on, still working fine and still seems to fix the issue. Any reason not to commit it? -- Matthew Fuller (MF4839) | fulle...@over-yonder.net Systems/Network Administrator | http://www.over-yonder.net/~fullermd/ On the Internet, nobody can hear you scream. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
On Mon, Mar 03, 2014 at 11:17:05AM +0200 I heard the voice of Andriy Gapon, and lo! it spake thus: I noticed that on some of our systems we were getting a clearly abnormal number of l2arc checksum errors accounted in l2_cksum_bad. [...] I propose the following patch which has been tested and seems to fix the problem without introducing any side effects: I've been running this for 2 weeks now without any cksum_bad's showing up (long enough that without it, I'd have expected them), and no other obvious side effects. FreeBSD 11.0-CURRENT #0 r264306M So if there are votes, I vote for making it work without that 'M' :) -- Matthew Fuller (MF4839) | fulle...@over-yonder.net Systems/Network Administrator | http://www.over-yonder.net/~fullermd/ On the Internet, nobody can hear you scream. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
On Mon, Mar 03, 2014 at 11:17:05AM +0200 I heard the voice of Andriy Gapon, and lo! it spake thus: I noticed that on some of our systems we were getting a clearly abnormal number of l2arc checksum errors accounted in l2_cksum_bad. The hardware appeared to be in good health. FWIW, I have a system here where I see similar things; after sufficient uptime it shows an impressive number and proportion of checksum errors (50% or more of the hits, eventually), since at least sometime last fall. No indication anywhere else of problems (CAM, SMART, zpool status). Haven't tried the patches. It would take most of a month of seeing nothing to have much confidence they did anything anyway; it's got 2 weeks now and only ~500 errors, so it takes a long time with this system/workload to ramp up. But it'd be a nice fix to have, so mark me up as a user-vote for landing if the code makes sense to the codesense people. -- Matthew Fuller (MF4839) | fulle...@over-yonder.net Systems/Network Administrator | http://www.over-yonder.net/~fullermd/ On the Internet, nobody can hear you scream. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
on 18/10/2013 17:57 Steven Hartland said the following: I think we we may well need the following patch to set the minblock size based on the vdev ashift and not SPA_MINBLOCKSIZE. svn diff -x -p sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c Index: sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c === --- sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c(revision 256554) +++ sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c(working copy) @@ -5147,7 +5147,7 @@ l2arc_compress_buf(l2arc_buf_hdr_t *l2hdr) len = l2hdr-b_asize; cdata = zio_data_buf_alloc(len); csize = zio_compress_data(ZIO_COMPRESS_LZ4, l2hdr-b_tmp_cdata, - cdata, l2hdr-b_asize, (size_t)SPA_MINBLOCKSIZE); + cdata, l2hdr-b_asize, (size_t)(1ULL l2hdr-b_dev-l2ad_vdev-vdev_ashift)); if (csize == 0) { /* zero block, indicate that there's nothing to write */ This is a rather old thread and change, but I think that I have identified another problem with 4KB cache devices. I noticed that on some of our systems we were getting a clearly abnormal number of l2arc checksum errors accounted in l2_cksum_bad. The hardware appeared to be in good health. Using DTrace I noticed that the data seemed to be overwritten with other data. After more DTrace analysis I observed that sometimes l2arc_write_buffers() would advance l2ad_hand by more than target_sz. This meant that l2arc_write_buffers() would write beyond a region cleared by l2arc_evict() and thus overwrite data belonging to non-evicted buffers. Havoc ensues. The cache devices in question are all SSDs with logical sector size of 4KB. I am not sure about other ZFS platforms, but on FreeBSD this fact is detected and ashift of 12 is used for the cache vdevs. Looking at l2arc_write_buffers() code you can see that it properly accounts for ashift when actually writing buffers and advancing l2ad_hand: /* * Keep the clock hand suitably device-aligned. */ buf_p_sz = vdev_psize_to_asize(dev-l2ad_vdev, buf_sz); write_psize += buf_p_sz; dev-l2ad_hand += buf_p_sz; But the same is not done when selecting buffers to be written and checking that target_sz is not exceeded. So, if ARC contains a lot of buffers smaller than 4K that means that an aligned on-disk size of the L2ARC buffers could be quite larger than their non-aligned size. I propose the following patch which has been tested and seems to fix the problem without introducing any side effects: https://github.com/avg-I/freebsd/compare/review;l2arc-write-target-size.diff https://github.com/avg-I/freebsd/compare/review;l2arc-write-target-size -- Andriy Gapon ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
- Original Message - From: Andriy Gapon a...@freebsd.org on 18/10/2013 17:57 Steven Hartland said the following: I think we we may well need the following patch to set the minblock size based on the vdev ashift and not SPA_MINBLOCKSIZE. svn diff -x -p sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c Index: sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c === --- sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c(revision 256554) +++ sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c(working copy) @@ -5147,7 +5147,7 @@ l2arc_compress_buf(l2arc_buf_hdr_t *l2hdr) len = l2hdr-b_asize; cdata = zio_data_buf_alloc(len); csize = zio_compress_data(ZIO_COMPRESS_LZ4, l2hdr-b_tmp_cdata, - cdata, l2hdr-b_asize, (size_t)SPA_MINBLOCKSIZE); + cdata, l2hdr-b_asize, (size_t)(1ULL l2hdr-b_dev-l2ad_vdev-vdev_ashift)); if (csize == 0) { /* zero block, indicate that there's nothing to write */ This is a rather old thread and change, but I think that I have identified another problem with 4KB cache devices. I noticed that on some of our systems we were getting a clearly abnormal number of l2arc checksum errors accounted in l2_cksum_bad. The hardware appeared to be in good health. Using DTrace I noticed that the data seemed to be overwritten with other data. After more DTrace analysis I observed that sometimes l2arc_write_buffers() would advance l2ad_hand by more than target_sz. This meant that l2arc_write_buffers() would write beyond a region cleared by l2arc_evict() and thus overwrite data belonging to non-evicted buffers. Havoc ensues. The cache devices in question are all SSDs with logical sector size of 4KB. I am not sure about other ZFS platforms, but on FreeBSD this fact is detected and ashift of 12 is used for the cache vdevs. Looking at l2arc_write_buffers() code you can see that it properly accounts for ashift when actually writing buffers and advancing l2ad_hand: /* * Keep the clock hand suitably device-aligned. */ buf_p_sz = vdev_psize_to_asize(dev-l2ad_vdev, buf_sz); write_psize += buf_p_sz; dev-l2ad_hand += buf_p_sz; But the same is not done when selecting buffers to be written and checking that target_sz is not exceeded. So, if ARC contains a lot of buffers smaller than 4K that means that an aligned on-disk size of the L2ARC buffers could be quite larger than their non-aligned size. I propose the following patch which has been tested and seems to fix the problem without introducing any side effects: https://github.com/avg-I/freebsd/compare/review;l2arc-write-target-size.diff https://github.com/avg-I/freebsd/compare/review;l2arc-write-target-size Looks good to me. Regards Steve This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Ок, just up to now no error on l2arc L2 ARC Summary: (HEALTHY) Passed Headroom:1.99m Tried Lock Failures:144.53m IO In Progress: 130.15k Low Memory Aborts: 7 Free on Write: 335.56k Writes While Full: 30.31k R/W Clashes:115.31k Bad Checksums: 0 IO Errors: 0 SPA Mismatch: 153.15m L2 ARC Size: (Adaptive) 433.75 GiB Header Size:0.49% 2.12GiB I will test for longer time, but looks like problem gone. Vitalij Satanivskij wrote: VS Steven Hartland wrote: VS SH So previously you only started seeing l2 errors after there was VS SH a significant amount of data in l2arc? Thats interesting in itself VS SH if thats the case. VS VS Yes someting arround 200+gb VS VS SH I wonder if its the type of data, or something similar. Do you VS SH run compression on any of your volumes? VS SH zfs get compression VS VS Just now testing goes on next configuration VS VS first zfs is top level pool calling disk1 have enable lz4 compression and secondarycache = metadata VS VS next zfs is disk1/data with compression=off and secondarycache = all VS VS Error was seen on confiruration like that and on configuration where was seted as secondarycache = none for disk1 (disk1/data still fully cached) VS VS VS VS VS SH Regards VS SH Steve VS SH - Original Message - VS SH From: Vitalij Satanivskij sa...@ukr.net VS SH VS SH VS SH VS SH Just now I cannot say, as to triger problem we need at last 200+gb size on l2arc wich usually grow in one production day. VS SH VS SH But for some reason today in the morning server was rebooted so cache was flushed and now only 100Gb. VS SH VS SH Need to wait some more time. VS SH VS SH At last for now none error on l2. VS SH VS SH VS SH VS SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. VS SH VS SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 VS SH or return the E.mail to postmas...@multiplay.co.uk. VS SH VS SH ___ VS SH freebsd-current@freebsd.org mailing list VS SH http://lists.freebsd.org/mailman/listinfo/freebsd-current VS SH To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org VS ___ VS freebsd-current@freebsd.org mailing list VS http://lists.freebsd.org/mailman/listinfo/freebsd-current VS To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Hows things looking Vitalij? - Original Message - From: Vitalij Satanivskij sa...@ukr.net Ok. Just right now system rebooted with you patch. Trim enabled again. WIll wait some time untile size of used cache grow's. Steven Hartland wrote: SH Looking at the l2arc compression code I believe that metadata is always SH compressed with lz4, even if compression is off on all datasets. SH SH This is backed up by what I'm seeing on my system here as it shows a SH non-zero l2_compress_successes value even though I'm not using SH compression at all. SH SH I think we we may well need the following patch to set the minblock SH size based on the vdev ashift and not SPA_MINBLOCKSIZE. SH SH svn diff -x -p sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c SH Index: sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c SH === SH --- sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c(revision 256554) SH +++ sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c(working copy) SH @@ -5147,7 +5147,7 @@ l2arc_compress_buf(l2arc_buf_hdr_t *l2hdr) SH len = l2hdr-b_asize; SH cdata = zio_data_buf_alloc(len); SH csize = zio_compress_data(ZIO_COMPRESS_LZ4, l2hdr-b_tmp_cdata, SH - cdata, l2hdr-b_asize, (size_t)SPA_MINBLOCKSIZE); SH + cdata, l2hdr-b_asize, (size_t)(1ULL l2hdr-b_dev-l2ad_vdev-vdev_ashift)); SH SH if (csize == 0) { SH /* zero block, indicate that there's nothing to write */ SH SH Could you try this patch on your system Vitalij see if it has any effect SH on the number of l2_cksum_bad / l2_io_error? This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Just now I cannot say, as to triger problem we need at last 200+gb size on l2arc wich usually grow in one production day. But for some reason today in the morning server was rebooted so cache was flushed and now only 100Gb. Need to wait some more time. At last for now none error on l2. Steven Hartland wrote: SH Hows things looking Vitalij? SH SH - Original Message - SH From: Vitalij Satanivskij sa...@ukr.net SH SH SH Ok. Just right now system rebooted with you patch. SH SH Trim enabled again. SH SH WIll wait some time untile size of used cache grow's. SH SH SH Steven Hartland wrote: SH SH Looking at the l2arc compression code I believe that metadata is always SH SH compressed with lz4, even if compression is off on all datasets. SH SH SH SH This is backed up by what I'm seeing on my system here as it shows a SH SH non-zero l2_compress_successes value even though I'm not using SH SH compression at all. SH SH SH SH I think we we may well need the following patch to set the minblock SH SH size based on the vdev ashift and not SPA_MINBLOCKSIZE. SH SH SH SH svn diff -x -p sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c SH SH Index: sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c SH SH === SH SH --- sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c (revision 256554) SH SH +++ sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c (working copy) SH SH @@ -5147,7 +5147,7 @@ l2arc_compress_buf(l2arc_buf_hdr_t *l2hdr) SH SH len = l2hdr-b_asize; SH SH cdata = zio_data_buf_alloc(len); SH SH csize = zio_compress_data(ZIO_COMPRESS_LZ4, l2hdr-b_tmp_cdata, SH SH - cdata, l2hdr-b_asize, (size_t)SPA_MINBLOCKSIZE); SH SH + cdata, l2hdr-b_asize, (size_t)(1ULL l2hdr-b_dev-l2ad_vdev-vdev_ashift)); SH SH SH SH if (csize == 0) { SH SH /* zero block, indicate that there's nothing to write */ SH SH SH SH Could you try this patch on your system Vitalij see if it has any effect SH SH on the number of l2_cksum_bad / l2_io_error? SH SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 SH or return the E.mail to postmas...@multiplay.co.uk. SH SH ___ SH freebsd-current@freebsd.org mailing list SH http://lists.freebsd.org/mailman/listinfo/freebsd-current SH To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
So previously you only started seeing l2 errors after there was a significant amount of data in l2arc? Thats interesting in itself if thats the case. I wonder if its the type of data, or something similar. Do you run compression on any of your volumes? zfs get compression Regards Steve - Original Message - From: Vitalij Satanivskij sa...@ukr.net Just now I cannot say, as to triger problem we need at last 200+gb size on l2arc wich usually grow in one production day. But for some reason today in the morning server was rebooted so cache was flushed and now only 100Gb. Need to wait some more time. At last for now none error on l2. This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Steven Hartland wrote: SH So previously you only started seeing l2 errors after there was SH a significant amount of data in l2arc? Thats interesting in itself SH if thats the case. Yes someting arround 200+gb SH I wonder if its the type of data, or something similar. Do you SH run compression on any of your volumes? SH zfs get compression Just now testing goes on next configuration first zfs is top level pool calling disk1 have enable lz4 compression and secondarycache = metadata next zfs is disk1/data with compression=off and secondarycache = all Error was seen on confiruration like that and on configuration where was seted as secondarycache = none for disk1 (disk1/data still fully cached) SH Regards SH Steve SH - Original Message - SH From: Vitalij Satanivskij sa...@ukr.net SH SH SH SH Just now I cannot say, as to triger problem we need at last 200+gb size on l2arc wich usually grow in one production day. SH SH But for some reason today in the morning server was rebooted so cache was flushed and now only 100Gb. SH SH Need to wait some more time. SH SH At last for now none error on l2. SH SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 SH or return the E.mail to postmas...@multiplay.co.uk. SH SH ___ SH freebsd-current@freebsd.org mailing list SH http://lists.freebsd.org/mailman/listinfo/freebsd-current SH To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Ok. Just right now system rebooted with you patch. Trim enabled again. WIll wait some time untile size of used cache grow's. Steven Hartland wrote: SH Looking at the l2arc compression code I believe that metadata is always SH compressed with lz4, even if compression is off on all datasets. SH SH This is backed up by what I'm seeing on my system here as it shows a SH non-zero l2_compress_successes value even though I'm not using SH compression at all. SH SH I think we we may well need the following patch to set the minblock SH size based on the vdev ashift and not SPA_MINBLOCKSIZE. SH SH svn diff -x -p sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c SH Index: sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c SH === SH --- sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c(revision 256554) SH +++ sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c(working copy) SH @@ -5147,7 +5147,7 @@ l2arc_compress_buf(l2arc_buf_hdr_t *l2hdr) SH len = l2hdr-b_asize; SH cdata = zio_data_buf_alloc(len); SH csize = zio_compress_data(ZIO_COMPRESS_LZ4, l2hdr-b_tmp_cdata, SH - cdata, l2hdr-b_asize, (size_t)SPA_MINBLOCKSIZE); SH + cdata, l2hdr-b_asize, (size_t)(1ULL l2hdr-b_dev-l2ad_vdev-vdev_ashift)); SH SH if (csize == 0) { SH /* zero block, indicate that there's nothing to write */ SH SH Could you try this patch on your system Vitalij see if it has any effect SH on the number of l2_cksum_bad / l2_io_error? SH SH Regards SH Steve SH - Original Message - SH From: Vitalij Satanivskij sa...@ukr.net SH To: Steven Hartland kill...@multiplay.co.uk SH Cc: Vitalij Satanivskij sa...@ukr.net; Dmitriy Makarov suppor...@ukr.net; Justin T. Gibbs gi...@freebsd.org; Borja SH Marcos bor...@sarenet.es; freebsd-current@freebsd.org SH Sent: Friday, October 18, 2013 3:45 PM SH Subject: Re: ZFS secondarycache on SSD problem on r255173 SH SH SH SH Just right now stats not to actual because of some another test. SH SH Test is simply all gpart information destroyed from ssd and SH SH They used as raw cache devices. Just SH 2013-10-18.11:30:49 zpool add disk1 cache /dev/ada1 /dev/ada2 /dev/ada3 SH SH So sizes at last l2_size and l2_asize in not actual. SH SH But heare it is: SH SH kstat.zfs.misc.arcstats.hits: 5178174063 SH kstat.zfs.misc.arcstats.misses: 57690806 SH kstat.zfs.misc.arcstats.demand_data_hits: 313995744 SH kstat.zfs.misc.arcstats.demand_data_misses: 37414740 SH kstat.zfs.misc.arcstats.demand_metadata_hits: 4719242892 SH kstat.zfs.misc.arcstats.demand_metadata_misses: 9266394 SH kstat.zfs.misc.arcstats.prefetch_data_hits: 1182495 SH kstat.zfs.misc.arcstats.prefetch_data_misses: 9951733 SH kstat.zfs.misc.arcstats.prefetch_metadata_hits: 143752935 SH kstat.zfs.misc.arcstats.prefetch_metadata_misses: 1057939 SH kstat.zfs.misc.arcstats.mru_hits: 118609738 SH kstat.zfs.misc.arcstats.mru_ghost_hits: 1895486 SH kstat.zfs.misc.arcstats.mfu_hits: 4914673425 SH kstat.zfs.misc.arcstats.mfu_ghost_hits: 14537497 SH kstat.zfs.misc.arcstats.allocated: 103796455 SH kstat.zfs.misc.arcstats.deleted: 40168100 SH kstat.zfs.misc.arcstats.stolen: 20832742 SH kstat.zfs.misc.arcstats.recycle_miss: 15663428 SH kstat.zfs.misc.arcstats.mutex_miss: 1456781 SH kstat.zfs.misc.arcstats.evict_skip: 25960184 SH kstat.zfs.misc.arcstats.evict_l2_cached: 891379153920 SH kstat.zfs.misc.arcstats.evict_l2_eligible: 50578438144 SH kstat.zfs.misc.arcstats.evict_l2_ineligible: 956055729664 SH kstat.zfs.misc.arcstats.hash_elements: 8693451 SH kstat.zfs.misc.arcstats.hash_elements_max: 14369414 SH kstat.zfs.misc.arcstats.hash_collisions: 90967764 SH kstat.zfs.misc.arcstats.hash_chains: 1891463 SH kstat.zfs.misc.arcstats.hash_chain_max: 24 SH kstat.zfs.misc.arcstats.p: 73170954752 SH kstat.zfs.misc.arcstats.c: 85899345920 SH kstat.zfs.misc.arcstats.c_min: 42949672960 SH kstat.zfs.misc.arcstats.c_max: 85899345920 SH kstat.zfs.misc.arcstats.size: 85899263104 SH kstat.zfs.misc.arcstats.hdr_size: 1425948696 SH kstat.zfs.misc.arcstats.data_size: 77769994240 SH kstat.zfs.misc.arcstats.other_size: 6056233632 SH kstat.zfs.misc.arcstats.l2_hits: 21725934 SH kstat.zfs.misc.arcstats.l2_misses: 35876251 SH kstat.zfs.misc.arcstats.l2_feeds: 130197 SH kstat.zfs.misc.arcstats.l2_rw_clash: 110181 SH kstat.zfs.misc.arcstats.l2_read_bytes: 391282009600 SH kstat.zfs.misc.arcstats.l2_write_bytes: 1098703347712 SH kstat.zfs.misc.arcstats.l2_writes_sent: 130037 SH kstat.zfs.misc.arcstats.l2_writes_done: 130037 SH kstat.zfs.misc.arcstats.l2_writes_error: 0 SH kstat.zfs.misc.arcstats.l2_writes_hdr_miss: 375921 SH kstat.zfs.misc.arcstats.l2_evict_lock_retry: 331 SH kstat.zfs.misc.arcstats.l2_evict_reading: 43 SH kstat.zfs.misc.arcstats.l2_free_on_write: 255730 SH kstat.zfs.misc.arcstats.l2_abort_lowmem: 0 SH kstat.zfs.misc.arcstats.l2_cksum_bad: 854359
Re: ZFS secondarycache on SSD problem on r255173
Hello. Yesterday system was rebooted with vfs.zfs.trim.enabled=0 System version 10.0-BETA1 FreeBSD 10.0-BETA1 #6 r256669, without any changes in code Uptime 10:51 up 16:41 sysctl vfs.zfs.trim.enabled vfs.zfs.trim.enabled: 0 Around 2 hours ago errors counter's kstat.zfs.misc.arcstats.l2_cksum_bad: 854359 kstat.zfs.misc.arcstats.l2_io_error: 38254 begin grow from zero values. After remove cache 2013-10-18.10:37:10 zpool remove disk1 gpt/cache0 gpt/cache1 gpt/cache2 and attach again 2013-10-18.10:38:28 zpool add disk1 cache gpt/cache0 gpt/cache1 gpt/cache2 counters stop growing (of couse thay not zeroed) before cache remove kstat.zfs.misc.arcstats.l2_asize was around 280GB hw size of l2 cache is 3x164G = 34 351651821 ada3 GPT (168G) 34 6- free - (3.0K) 408388608 1 zil2 (4.0G) 8388648 343263200 2 cache2 (164G) 351651848 7- free - (3.5K) Any hypothesis what alse we can test/try etc? Steven Hartland wrote: SH Correct. SH - Original Message - SH From: Vitalij Satanivskij sa...@ukr.net SH SH SH Just to be sure I understand you clearly, I need to test next configuration: SH SH 1) System with ashift patch eg. just latest stable/10 revision SH 2) vfs.zfs.trim.enabled=0 in /boot/loader.conf SH SH So realy only diferens in default system configuration is disabled trim functional ? SH SH SH SH Steven Hartland wrote: SH SH Still worth testing with the problem version installed but SH SH with trim disabled to see if that clears the issues, if SH SH nothing else it will confirm / deny if trim is involved. SH SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 SH or return the E.mail to postmas...@multiplay.co.uk. SH SH ___ SH freebsd-current@freebsd.org mailing list SH http://lists.freebsd.org/mailman/listinfo/freebsd-current SH To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Hmm so that rules out a TRIM related issue. I wonder if the increase in ashift has triggered a problem in compression. What are all the values reported by: sysctl -a kstat.zfs.misc.arcstats Regards Steve - Original Message - From: Vitalij Satanivskij sa...@ukr.net To: Steven Hartland kill...@multiplay.co.uk Cc: sa...@ukr.net; Justin T. Gibbs gi...@freebsd.org; freebsd-current@freebsd.org; Borja Marcos bor...@sarenet.es; Dmitriy Makarov suppor...@ukr.net Sent: Friday, October 18, 2013 9:01 AM Subject: Re: ZFS secondarycache on SSD problem on r255173 Hello. Yesterday system was rebooted with vfs.zfs.trim.enabled=0 System version 10.0-BETA1 FreeBSD 10.0-BETA1 #6 r256669, without any changes in code Uptime 10:51 up 16:41 sysctl vfs.zfs.trim.enabled vfs.zfs.trim.enabled: 0 Around 2 hours ago errors counter's kstat.zfs.misc.arcstats.l2_cksum_bad: 854359 kstat.zfs.misc.arcstats.l2_io_error: 38254 begin grow from zero values. After remove cache 2013-10-18.10:37:10 zpool remove disk1 gpt/cache0 gpt/cache1 gpt/cache2 and attach again 2013-10-18.10:38:28 zpool add disk1 cache gpt/cache0 gpt/cache1 gpt/cache2 counters stop growing (of couse thay not zeroed) before cache remove kstat.zfs.misc.arcstats.l2_asize was around 280GB hw size of l2 cache is 3x164G = 34 351651821 ada3 GPT (168G) 34 6- free - (3.0K) 408388608 1 zil2 (4.0G) 8388648 343263200 2 cache2 (164G) 351651848 7- free - (3.5K) Any hypothesis what alse we can test/try etc? Steven Hartland wrote: SH Correct. SH - Original Message - SH From: Vitalij Satanivskij sa...@ukr.net SH SH SH Just to be sure I understand you clearly, I need to test next configuration: SH SH 1) System with ashift patch eg. just latest stable/10 revision SH 2) vfs.zfs.trim.enabled=0 in /boot/loader.conf SH SH So realy only diferens in default system configuration is disabled trim functional ? SH SH SH SH Steven Hartland wrote: SH SH Still worth testing with the problem version installed but SH SH with trim disabled to see if that clears the issues, if SH SH nothing else it will confirm / deny if trim is involved. SH SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 SH or return the E.mail to postmas...@multiplay.co.uk. SH SH ___ SH freebsd-current@freebsd.org mailing list SH http://lists.freebsd.org/mailman/listinfo/freebsd-current SH To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Just right now stats not to actual because of some another test. Test is simply all gpart information destroyed from ssd and They used as raw cache devices. Just 2013-10-18.11:30:49 zpool add disk1 cache /dev/ada1 /dev/ada2 /dev/ada3 So sizes at last l2_size and l2_asize in not actual. But heare it is: kstat.zfs.misc.arcstats.hits: 5178174063 kstat.zfs.misc.arcstats.misses: 57690806 kstat.zfs.misc.arcstats.demand_data_hits: 313995744 kstat.zfs.misc.arcstats.demand_data_misses: 37414740 kstat.zfs.misc.arcstats.demand_metadata_hits: 4719242892 kstat.zfs.misc.arcstats.demand_metadata_misses: 9266394 kstat.zfs.misc.arcstats.prefetch_data_hits: 1182495 kstat.zfs.misc.arcstats.prefetch_data_misses: 9951733 kstat.zfs.misc.arcstats.prefetch_metadata_hits: 143752935 kstat.zfs.misc.arcstats.prefetch_metadata_misses: 1057939 kstat.zfs.misc.arcstats.mru_hits: 118609738 kstat.zfs.misc.arcstats.mru_ghost_hits: 1895486 kstat.zfs.misc.arcstats.mfu_hits: 4914673425 kstat.zfs.misc.arcstats.mfu_ghost_hits: 14537497 kstat.zfs.misc.arcstats.allocated: 103796455 kstat.zfs.misc.arcstats.deleted: 40168100 kstat.zfs.misc.arcstats.stolen: 20832742 kstat.zfs.misc.arcstats.recycle_miss: 15663428 kstat.zfs.misc.arcstats.mutex_miss: 1456781 kstat.zfs.misc.arcstats.evict_skip: 25960184 kstat.zfs.misc.arcstats.evict_l2_cached: 891379153920 kstat.zfs.misc.arcstats.evict_l2_eligible: 50578438144 kstat.zfs.misc.arcstats.evict_l2_ineligible: 956055729664 kstat.zfs.misc.arcstats.hash_elements: 8693451 kstat.zfs.misc.arcstats.hash_elements_max: 14369414 kstat.zfs.misc.arcstats.hash_collisions: 90967764 kstat.zfs.misc.arcstats.hash_chains: 1891463 kstat.zfs.misc.arcstats.hash_chain_max: 24 kstat.zfs.misc.arcstats.p: 73170954752 kstat.zfs.misc.arcstats.c: 85899345920 kstat.zfs.misc.arcstats.c_min: 42949672960 kstat.zfs.misc.arcstats.c_max: 85899345920 kstat.zfs.misc.arcstats.size: 85899263104 kstat.zfs.misc.arcstats.hdr_size: 1425948696 kstat.zfs.misc.arcstats.data_size: 77769994240 kstat.zfs.misc.arcstats.other_size: 6056233632 kstat.zfs.misc.arcstats.l2_hits: 21725934 kstat.zfs.misc.arcstats.l2_misses: 35876251 kstat.zfs.misc.arcstats.l2_feeds: 130197 kstat.zfs.misc.arcstats.l2_rw_clash: 110181 kstat.zfs.misc.arcstats.l2_read_bytes: 391282009600 kstat.zfs.misc.arcstats.l2_write_bytes: 1098703347712 kstat.zfs.misc.arcstats.l2_writes_sent: 130037 kstat.zfs.misc.arcstats.l2_writes_done: 130037 kstat.zfs.misc.arcstats.l2_writes_error: 0 kstat.zfs.misc.arcstats.l2_writes_hdr_miss: 375921 kstat.zfs.misc.arcstats.l2_evict_lock_retry: 331 kstat.zfs.misc.arcstats.l2_evict_reading: 43 kstat.zfs.misc.arcstats.l2_free_on_write: 255730 kstat.zfs.misc.arcstats.l2_abort_lowmem: 0 kstat.zfs.misc.arcstats.l2_cksum_bad: 854359 kstat.zfs.misc.arcstats.l2_io_error: 38254 kstat.zfs.misc.arcstats.l2_size: 136696884736 kstat.zfs.misc.arcstats.l2_asize: 131427690496 kstat.zfs.misc.arcstats.l2_hdr_size: 742951208 kstat.zfs.misc.arcstats.l2_compress_successes: 5565311 kstat.zfs.misc.arcstats.l2_compress_zeros: 0 kstat.zfs.misc.arcstats.l2_compress_failures: 0 kstat.zfs.misc.arcstats.l2_write_trylock_fail: 325157131 kstat.zfs.misc.arcstats.l2_write_passed_headroom: 4897854 kstat.zfs.misc.arcstats.l2_write_spa_mismatch: 115704249 kstat.zfs.misc.arcstats.l2_write_in_l2: 15114214372 kstat.zfs.misc.arcstats.l2_write_io_in_progress: 63417 kstat.zfs.misc.arcstats.l2_write_not_cacheable: 3291593934 kstat.zfs.misc.arcstats.l2_write_full: 47672 kstat.zfs.misc.arcstats.l2_write_buffer_iter: 130197 kstat.zfs.misc.arcstats.l2_write_pios: 130037 kstat.zfs.misc.arcstats.l2_write_buffer_bytes_scanned: 369077156457472 kstat.zfs.misc.arcstats.l2_write_buffer_list_iter: 8015080 kstat.zfs.misc.arcstats.l2_write_buffer_list_null_iter: 79825 kstat.zfs.misc.arcstats.memory_throttle_count: 0 kstat.zfs.misc.arcstats.duplicate_buffers: 0 kstat.zfs.misc.arcstats.duplicate_buffers_size: 0 kstat.zfs.misc.arcstats.duplicate_reads: 0 Values of - kstat.zfs.misc.arcstats.l2_cksum_bad: 854359 kstat.zfs.misc.arcstats.l2_io_error: 38254 not growing from last cache reconfiguration, just wait some time to see - maybe problem disapers :) Steven Hartland wrote: SH Hmm so that rules out a TRIM related issue. I wonder if the SH increase in ashift has triggered a problem in compression. SH SH What are all the values reported by: SH sysctl -a kstat.zfs.misc.arcstats SH SH Regards SH Steve SH SH - Original Message - SH From: Vitalij Satanivskij sa...@ukr.net SH To: Steven Hartland kill...@multiplay.co.uk SH Cc: sa...@ukr.net; Justin T. Gibbs gi...@freebsd.org; freebsd-current@freebsd.org; Borja Marcos bor...@sarenet.es; SH Dmitriy Makarov suppor...@ukr.net SH Sent: Friday, October 18, 2013 9:01 AM SH Subject: Re: ZFS secondarycache on SSD problem on r255173 SH SH SH Hello. SH SH Yesterday system was rebooted with vfs.zfs.trim.enabled=0 SH SH System version 10.0-BETA1 FreeBSD 10.0-BETA1 #6
Re: ZFS secondarycache on SSD problem on r255173
Looking at the l2arc compression code I believe that metadata is always compressed with lz4, even if compression is off on all datasets. This is backed up by what I'm seeing on my system here as it shows a non-zero l2_compress_successes value even though I'm not using compression at all. I think we we may well need the following patch to set the minblock size based on the vdev ashift and not SPA_MINBLOCKSIZE. svn diff -x -p sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c Index: sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c === --- sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c(revision 256554) +++ sys/cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c(working copy) @@ -5147,7 +5147,7 @@ l2arc_compress_buf(l2arc_buf_hdr_t *l2hdr) len = l2hdr-b_asize; cdata = zio_data_buf_alloc(len); csize = zio_compress_data(ZIO_COMPRESS_LZ4, l2hdr-b_tmp_cdata, - cdata, l2hdr-b_asize, (size_t)SPA_MINBLOCKSIZE); + cdata, l2hdr-b_asize, (size_t)(1ULL l2hdr-b_dev-l2ad_vdev-vdev_ashift)); if (csize == 0) { /* zero block, indicate that there's nothing to write */ Could you try this patch on your system Vitalij see if it has any effect on the number of l2_cksum_bad / l2_io_error? Regards Steve - Original Message - From: Vitalij Satanivskij sa...@ukr.net To: Steven Hartland kill...@multiplay.co.uk Cc: Vitalij Satanivskij sa...@ukr.net; Dmitriy Makarov suppor...@ukr.net; Justin T. Gibbs gi...@freebsd.org; Borja Marcos bor...@sarenet.es; freebsd-current@freebsd.org Sent: Friday, October 18, 2013 3:45 PM Subject: Re: ZFS secondarycache on SSD problem on r255173 Just right now stats not to actual because of some another test. Test is simply all gpart information destroyed from ssd and They used as raw cache devices. Just 2013-10-18.11:30:49 zpool add disk1 cache /dev/ada1 /dev/ada2 /dev/ada3 So sizes at last l2_size and l2_asize in not actual. But heare it is: kstat.zfs.misc.arcstats.hits: 5178174063 kstat.zfs.misc.arcstats.misses: 57690806 kstat.zfs.misc.arcstats.demand_data_hits: 313995744 kstat.zfs.misc.arcstats.demand_data_misses: 37414740 kstat.zfs.misc.arcstats.demand_metadata_hits: 4719242892 kstat.zfs.misc.arcstats.demand_metadata_misses: 9266394 kstat.zfs.misc.arcstats.prefetch_data_hits: 1182495 kstat.zfs.misc.arcstats.prefetch_data_misses: 9951733 kstat.zfs.misc.arcstats.prefetch_metadata_hits: 143752935 kstat.zfs.misc.arcstats.prefetch_metadata_misses: 1057939 kstat.zfs.misc.arcstats.mru_hits: 118609738 kstat.zfs.misc.arcstats.mru_ghost_hits: 1895486 kstat.zfs.misc.arcstats.mfu_hits: 4914673425 kstat.zfs.misc.arcstats.mfu_ghost_hits: 14537497 kstat.zfs.misc.arcstats.allocated: 103796455 kstat.zfs.misc.arcstats.deleted: 40168100 kstat.zfs.misc.arcstats.stolen: 20832742 kstat.zfs.misc.arcstats.recycle_miss: 15663428 kstat.zfs.misc.arcstats.mutex_miss: 1456781 kstat.zfs.misc.arcstats.evict_skip: 25960184 kstat.zfs.misc.arcstats.evict_l2_cached: 891379153920 kstat.zfs.misc.arcstats.evict_l2_eligible: 50578438144 kstat.zfs.misc.arcstats.evict_l2_ineligible: 956055729664 kstat.zfs.misc.arcstats.hash_elements: 8693451 kstat.zfs.misc.arcstats.hash_elements_max: 14369414 kstat.zfs.misc.arcstats.hash_collisions: 90967764 kstat.zfs.misc.arcstats.hash_chains: 1891463 kstat.zfs.misc.arcstats.hash_chain_max: 24 kstat.zfs.misc.arcstats.p: 73170954752 kstat.zfs.misc.arcstats.c: 85899345920 kstat.zfs.misc.arcstats.c_min: 42949672960 kstat.zfs.misc.arcstats.c_max: 85899345920 kstat.zfs.misc.arcstats.size: 85899263104 kstat.zfs.misc.arcstats.hdr_size: 1425948696 kstat.zfs.misc.arcstats.data_size: 77769994240 kstat.zfs.misc.arcstats.other_size: 6056233632 kstat.zfs.misc.arcstats.l2_hits: 21725934 kstat.zfs.misc.arcstats.l2_misses: 35876251 kstat.zfs.misc.arcstats.l2_feeds: 130197 kstat.zfs.misc.arcstats.l2_rw_clash: 110181 kstat.zfs.misc.arcstats.l2_read_bytes: 391282009600 kstat.zfs.misc.arcstats.l2_write_bytes: 1098703347712 kstat.zfs.misc.arcstats.l2_writes_sent: 130037 kstat.zfs.misc.arcstats.l2_writes_done: 130037 kstat.zfs.misc.arcstats.l2_writes_error: 0 kstat.zfs.misc.arcstats.l2_writes_hdr_miss: 375921 kstat.zfs.misc.arcstats.l2_evict_lock_retry: 331 kstat.zfs.misc.arcstats.l2_evict_reading: 43 kstat.zfs.misc.arcstats.l2_free_on_write: 255730 kstat.zfs.misc.arcstats.l2_abort_lowmem: 0 kstat.zfs.misc.arcstats.l2_cksum_bad: 854359 kstat.zfs.misc.arcstats.l2_io_error: 38254 kstat.zfs.misc.arcstats.l2_size: 136696884736 kstat.zfs.misc.arcstats.l2_asize: 131427690496 kstat.zfs.misc.arcstats.l2_hdr_size: 742951208 kstat.zfs.misc.arcstats.l2_compress_successes: 5565311 kstat.zfs.misc.arcstats.l2_compress_zeros: 0 kstat.zfs.misc.arcstats.l2_compress_failures: 0 kstat.zfs.misc.arcstats.l2_write_trylock_fail: 325157131 kstat.zfs.misc.arcstats.l2_write_passed_headroom: 4897854 kstat.zfs.misc.arcstats.l2_write_spa_mismatch: 115704249
Re: ZFS secondarycache on SSD problem on r255173
Hello. Problem description is in - http://lists.freebsd.org/pipermail/freebsd-current/2013-October/045088.html As we find later first begin's problem with errors counter in arcstats than size of l2 grows abnormal. After patch rollback everything is ok. Justin T. Gibbs wrote: JTG You'll have to be more specific. I don't have that email or know what list on which to search. JTG JTG Thanks, JTG Justin JTG JTG On Oct 16, 2013, at 2:01 AM, Vitalij Satanivskij sa...@ukr.net wrote: JTG JTG Hello. JTG JTG Patch brocke cache functionality. JTG JTG Look at's Dmitriy's mail from Mon, 07 Oct 2013 21:09:06 +0300 JTG JTG With subject ZFS L2ARC - incorrect size and abnormal system load on r255173 JTG JTG As patch alredy in head and BETA it's not good. JTG JTG Yesterday we update one machine up to beta1 and forgot about patch. So 12 Hours and cache broken... :(( JTG JTG JTG JTG Dmitriy Makarov wrote: JTG DM The attached patch by Steven Hartland fixes issue for me too. Thank you! JTG DM JTG DM JTG DM --- Исходное сообщение --- JTG DM От кого: Steven Hartland kill...@multiplay.co.uk JTG DM Дата: 18 сентября 2013, 01:53:10 JTG DM JTG DM - Original Message - JTG DM From: Justin T. Gibbs JTG DM JTG DM --- JTG DM Дмитрий Макаров JTG DM ___ JTG DM freebsd-current@freebsd.org mailing list JTG DM http://lists.freebsd.org/mailman/listinfo/freebsd-current JTG DM To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org JTG JTG JTG ___ JTG freebsd-current@freebsd.org mailing list JTG http://lists.freebsd.org/mailman/listinfo/freebsd-current JTG To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Hello. SSD is Intel SSD 530 series (INTEL SSDSC2BW180A4 DC12) Controler is onboard intel sata controler, motherboard is Supermicro X9SRL-F so it's Intel C602 chipset All cache ssd connected to sata 2 ports. System has LSI MPS controler (SAS2308) with firmware version - 16.00.00.00, but only hdd's (36 1TB WD RE4 drives) connected to it. Steven Hartland wrote: SH Ohh stupid question what hardware are you running this on, SH specifically what SSD's and what controller and if relavent SH what controller Firmware version? SH SH I wonder if you might have bad HW / FW, such as older LSI SH mps Firmware, which is know to causing corruption with SH some delete methods. SH SH Without the ashift fixes, you'll be sending short requests SH to the disk which will then ignore them, so I can see a SH potential corrilation there. SH SH You could rule this out by disabling ZFS TRIM with SH vfs.zfs.trim.enabled=0 in /boot/loader.conf SH SH Regards SH Steve SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 SH or return the E.mail to postmas...@multiplay.co.uk. SH SH ___ SH freebsd-current@freebsd.org mailing list SH http://lists.freebsd.org/mailman/listinfo/freebsd-current SH To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Still worth testing with the problem version installed but with trim disabled to see if that clears the issues, if nothing else it will confirm / deny if trim is involved. Regards Steve - Original Message - From: Vitalij Satanivskij sa...@ukr.net To: Steven Hartland kill...@multiplay.co.uk Cc: Justin T. Gibbs gi...@freebsd.org; Vitalij Satanivskij sa...@ukr.net; freebsd-current@freebsd.org; Borja Marcos bor...@sarenet.es; Dmitriy Makarov suppor...@ukr.net Sent: Thursday, October 17, 2013 7:12 AM Subject: Re: ZFS secondarycache on SSD problem on r255173 Hello. SSD is Intel SSD 530 series (INTEL SSDSC2BW180A4 DC12) Controler is onboard intel sata controler, motherboard is Supermicro X9SRL-F so it's Intel C602 chipset All cache ssd connected to sata 2 ports. System has LSI MPS controler (SAS2308) with firmware version - 16.00.00.00, but only hdd's (36 1TB WD RE4 drives) connected to it. Steven Hartland wrote: SH Ohh stupid question what hardware are you running this on, SH specifically what SSD's and what controller and if relavent SH what controller Firmware version? SH SH I wonder if you might have bad HW / FW, such as older LSI SH mps Firmware, which is know to causing corruption with SH some delete methods. SH SH Without the ashift fixes, you'll be sending short requests SH to the disk which will then ignore them, so I can see a SH potential corrilation there. SH SH You could rule this out by disabling ZFS TRIM with SH vfs.zfs.trim.enabled=0 in /boot/loader.conf SH SH Regards SH Steve SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 SH or return the E.mail to postmas...@multiplay.co.uk. SH SH ___ SH freebsd-current@freebsd.org mailing list SH http://lists.freebsd.org/mailman/listinfo/freebsd-current SH To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Just to be sure I understand you clearly, I need to test next configuration: 1) System with ashift patch eg. just latest stable/10 revision 2) vfs.zfs.trim.enabled=0 in /boot/loader.conf So realy only diferens in default system configuration is disabled trim functional ? Steven Hartland wrote: SH Still worth testing with the problem version installed but SH with trim disabled to see if that clears the issues, if SH nothing else it will confirm / deny if trim is involved. SH SH Regards SH Steve SH SH - Original Message - SH From: Vitalij Satanivskij sa...@ukr.net SH To: Steven Hartland kill...@multiplay.co.uk SH Cc: Justin T. Gibbs gi...@freebsd.org; Vitalij Satanivskij sa...@ukr.net; freebsd-current@freebsd.org; Borja Marcos SH bor...@sarenet.es; Dmitriy Makarov suppor...@ukr.net SH Sent: Thursday, October 17, 2013 7:12 AM SH Subject: Re: ZFS secondarycache on SSD problem on r255173 SH SH SH Hello. SH SH SSD is Intel SSD 530 series (INTEL SSDSC2BW180A4 DC12) SH SH Controler is onboard intel sata controler, motherboard is Supermicro X9SRL-F so it's Intel C602 chipset SH SH All cache ssd connected to sata 2 ports. SH SH System has LSI MPS controler (SAS2308) with firmware version - 16.00.00.00, but only hdd's (36 1TB WD RE4 drives) SH connected to it. SH SH SH Steven Hartland wrote: SH SH Ohh stupid question what hardware are you running this on, SH SH specifically what SSD's and what controller and if relavent SH SH what controller Firmware version? SH SH SH SH I wonder if you might have bad HW / FW, such as older LSI SH SH mps Firmware, which is know to causing corruption with SH SH some delete methods. SH SH SH SH Without the ashift fixes, you'll be sending short requests SH SH to the disk which will then ignore them, so I can see a SH SH potential corrilation there. SH SH SH SH You could rule this out by disabling ZFS TRIM with SH SH vfs.zfs.trim.enabled=0 in /boot/loader.conf SH SH SH SH Regards SH SH Steve SH SH SH SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the SH event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any SH information contained in it. SH SH SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 SH SH or return the E.mail to postmas...@multiplay.co.uk. SH SH SH SH ___ SH SH freebsd-current@freebsd.org mailing list SH SH http://lists.freebsd.org/mailman/listinfo/freebsd-current SH SH To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org SH SH SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 SH or return the E.mail to postmas...@multiplay.co.uk. SH ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Correct. - Original Message - From: Vitalij Satanivskij sa...@ukr.net Just to be sure I understand you clearly, I need to test next configuration: 1) System with ashift patch eg. just latest stable/10 revision 2) vfs.zfs.trim.enabled=0 in /boot/loader.conf So realy only diferens in default system configuration is disabled trim functional ? Steven Hartland wrote: SH Still worth testing with the problem version installed but SH with trim disabled to see if that clears the issues, if SH nothing else it will confirm / deny if trim is involved. This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Hello. Patch brocke cache functionality. Look at's Dmitriy's mail from Mon, 07 Oct 2013 21:09:06 +0300 With subject ZFS L2ARC - incorrect size and abnormal system load on r255173 As patch alredy in head and BETA it's not good. Yesterday we update one machine up to beta1 and forgot about patch. So 12 Hours and cache broken... :(( Dmitriy Makarov wrote: DM The attached patch by Steven Hartland fixes issue for me too. Thank you! DM DM DM --- Исходное сообщение --- DM От кого: Steven Hartland kill...@multiplay.co.uk DM Дата: 18 сентября 2013, 01:53:10 DM DM - Original Message - DM From: Justin T. Gibbs DM DM --- DM Дмитрий Макаров DM ___ DM freebsd-current@freebsd.org mailing list DM http://lists.freebsd.org/mailman/listinfo/freebsd-current DM To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Have you confirmed the ashift changes are the actual cause of this by backing out just those changes and retesting on the same hardware. Also worth checking your disks smart values to confirm there are no visible signs of HW errors. Regards Steve - Original Message - From: Vitalij Satanivskij sa...@ukr.net To: Dmitriy Makarov suppor...@ukr.net Cc: Steven Hartland kill...@multiplay.co.uk; Justin T. Gibbs gi...@freebsd.org; Borja Marcos bor...@sarenet.es; freebsd-current@freebsd.org Sent: Wednesday, October 16, 2013 9:01 AM Subject: Re: ZFS secondarycache on SSD problem on r255173 Hello. Patch brocke cache functionality. Look at's Dmitriy's mail from Mon, 07 Oct 2013 21:09:06 +0300 With subject ZFS L2ARC - incorrect size and abnormal system load on r255173 As patch alredy in head and BETA it's not good. Yesterday we update one machine up to beta1 and forgot about patch. So 12 Hours and cache broken... :(( Dmitriy Makarov wrote: DM The attached patch by Steven Hartland fixes issue for me too. Thank you! DM DM DM --- Исходное сообщение --- DM От кого: Steven Hartland kill...@multiplay.co.uk DM Дата: 18 сентября 2013, 01:53:10 DM DM - Original Message - DM From: Justin T. Gibbs DM DM --- DM Дмитрий Макаров DM ___ DM freebsd-current@freebsd.org mailing list DM http://lists.freebsd.org/mailman/listinfo/freebsd-current DM To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Yes We have 15 servers, all of them have problem while using with patch fo ashift, sh we rollback path (for r255173) and all of them works for a week without that's problem's. Yesterday one of of servers was updated to stable/10 (beta1) wich include patch and after around 12 hours of works l2arc begin et errors like that kstat.zfs.misc.arcstats.l2_cksum_bad kstat.zfs.misc.arcstats.l2_io_error For now patch disabled in ower production. Please note we have very heavy load on zfs pool so 90GB arc and 3x180Gb L2arc have very big hit's on it on it. SSD used for cache's is intel ssd 530 series smart for all devices in in normal states's no bad values on it. Steven Hartland wrote: SH Have you confirmed the ashift changes are the actual cause of this SH by backing out just those changes and retesting on the same hardware. SH SH Also worth checking your disks smart values to confirm there are no SH visible signs of HW errors. SH SH Regards SH Steve SH SH - Original Message - SH From: Vitalij Satanivskij sa...@ukr.net SH To: Dmitriy Makarov suppor...@ukr.net SH Cc: Steven Hartland kill...@multiplay.co.uk; Justin T. Gibbs gi...@freebsd.org; Borja Marcos bor...@sarenet.es; SH freebsd-current@freebsd.org SH Sent: Wednesday, October 16, 2013 9:01 AM SH Subject: Re: ZFS secondarycache on SSD problem on r255173 SH SH SH Hello. SH SH Patch brocke cache functionality. SH SH Look at's Dmitriy's mail from Mon, 07 Oct 2013 21:09:06 +0300 SH SH With subject ZFS L2ARC - incorrect size and abnormal system load on r255173 SH SH As patch alredy in head and BETA it's not good. SH SH Yesterday we update one machine up to beta1 and forgot about patch. So 12 Hours and cache broken... :(( SH SH SH SH Dmitriy Makarov wrote: SH DM The attached patch by Steven Hartland fixes issue for me too. Thank you! SH DM SH DM SH DM --- Исходное сообщение --- SH DM От кого: Steven Hartland kill...@multiplay.co.uk SH DM Дата: 18 сентября 2013, 01:53:10 SH DM SH DM - Original Message - SH DM From: Justin T. Gibbs SH DM SH DM --- SH DM Дмитрий Макаров SH DM ___ SH DM freebsd-current@freebsd.org mailing list SH DM http://lists.freebsd.org/mailman/listinfo/freebsd-current SH DM To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org SH SH SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 SH or return the E.mail to postmas...@multiplay.co.uk. SH SH ___ SH freebsd-current@freebsd.org mailing list SH http://lists.freebsd.org/mailman/listinfo/freebsd-current SH To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
I'm not clear what you rolled back there as r255173 has ntothing to do with this. Could you clarify Any errors recorded in /var/log/messages? Could you add code to record the none zero value of zio-io_error in l2arc_read_done as this may give some indication of the underlying issue. Additionally could always put a panic in that code path too and then create a dump so the details can be fully exhamined. In terms of the slowness thats going to be a side effect of the cache failures. Oh could you also confirm that the issue doesn't exist if you 1. Exclude r255753 2. Set vfs.zfs.max_auto_ashift=9 Regards Steve - Original Message - From: Vitalij Satanivskij sa...@ukr.net To: Steven Hartland kill...@multiplay.co.uk Cc: Vitalij Satanivskij sa...@ukr.net; Dmitriy Makarov suppor...@ukr.net; Justin T. Gibbs gi...@freebsd.org; Borja Marcos bor...@sarenet.es; freebsd-current@freebsd.org Sent: Wednesday, October 16, 2013 3:10 PM Subject: Re: ZFS secondarycache on SSD problem on r255173 Yes We have 15 servers, all of them have problem while using with patch fo ashift, sh we rollback path (for r255173) and all of them works for a week without that's problem's. Yesterday one of of servers was updated to stable/10 (beta1) wich include patch and after around 12 hours of works l2arc begin et errors like that kstat.zfs.misc.arcstats.l2_cksum_bad kstat.zfs.misc.arcstats.l2_io_error For now patch disabled in ower production. Please note we have very heavy load on zfs pool so 90GB arc and 3x180Gb L2arc have very big hit's on it on it. SSD used for cache's is intel ssd 530 series smart for all devices in in normal states's no bad values on it. Steven Hartland wrote: SH Have you confirmed the ashift changes are the actual cause of this SH by backing out just those changes and retesting on the same hardware. SH SH Also worth checking your disks smart values to confirm there are no SH visible signs of HW errors. SH SH Regards SH Steve SH SH - Original Message - SH From: Vitalij Satanivskij sa...@ukr.net SH To: Dmitriy Makarov suppor...@ukr.net SH Cc: Steven Hartland kill...@multiplay.co.uk; Justin T. Gibbs gi...@freebsd.org; Borja Marcos bor...@sarenet.es; SH freebsd-current@freebsd.org SH Sent: Wednesday, October 16, 2013 9:01 AM SH Subject: Re: ZFS secondarycache on SSD problem on r255173 SH SH SH Hello. SH SH Patch brocke cache functionality. SH SH Look at's Dmitriy's mail from Mon, 07 Oct 2013 21:09:06 +0300 SH SH With subject ZFS L2ARC - incorrect size and abnormal system load on r255173 SH SH As patch alredy in head and BETA it's not good. SH SH Yesterday we update one machine up to beta1 and forgot about patch. So 12 Hours and cache broken... :(( SH SH SH SH Dmitriy Makarov wrote: SH DM The attached patch by Steven Hartland fixes issue for me too. Thank you! SH DM SH DM SH DM --- Исходное сообщение --- SH DM От кого: Steven Hartland kill...@multiplay.co.uk SH DM Дата: 18 сентября 2013, 01:53:10 SH DM SH DM - Original Message - SH DM From: Justin T. Gibbs SH DM SH DM --- SH DM Дмитрий Макаров SH DM ___ SH DM freebsd-current@freebsd.org mailing list SH DM http://lists.freebsd.org/mailman/listinfo/freebsd-current SH DM To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org SH SH SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 SH or return the E.mail to postmas...@multiplay.co.uk. SH SH ___ SH freebsd-current@freebsd.org mailing list SH http://lists.freebsd.org/mailman/listinfo/freebsd-current SH To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Steven Hartland wrote: SH I'm not clear what you rolled back there as r255173 has ntothing to do SH with this. Could you clarify r255173 with you patch from email dated Tue, 17 Sep 2013 23:53:12 +0100 with subject Re: ZFS secondarycache on SSD problem on r255173 Errors wich we gets is in arcstats count not in messages, and was desribed some time ago in mails be me and Dmitriy Makarov with subject's ZFS L2ARC - incorrect size and abnormal system load on r255173 On r255173 without patch and with vfs.zfs.max_auto_ashift=9 when added to pool 2 ssd as caches get cache gpt/cache1ONLINE 0 0 0 block size: 512B configured, 4096B native gpt/cache2ONLINE 0 0 0 block size: 512B configured, 4096B native Same message we seen with default vfs.zfs.max_auto_ashift Will wait some time to see how it works. SH Any errors recorded in /var/log/messages? SH SH Could you add code to record the none zero value of zio-io_error in SH l2arc_read_done as this may give some indication of the underlying SH issue. SH SH Additionally could always put a panic in that code path too and then SH create a dump so the details can be fully exhamined. SH SH In terms of the slowness thats going to be a side effect of the cache SH failures. SH SH Oh could you also confirm that the issue doesn't exist if you SH 1. Exclude r255753 SH 2. Set vfs.zfs.max_auto_ashift=9 SH SH Regards SH Steve SH - Original Message - SH From: Vitalij Satanivskij sa...@ukr.net SH To: Steven Hartland kill...@multiplay.co.uk SH Cc: Vitalij Satanivskij sa...@ukr.net; Dmitriy Makarov suppor...@ukr.net; Justin T. Gibbs gi...@freebsd.org; Borja SH Marcos bor...@sarenet.es; freebsd-current@freebsd.org SH Sent: Wednesday, October 16, 2013 3:10 PM SH Subject: Re: ZFS secondarycache on SSD problem on r255173 SH SH SH Yes SH SH We have 15 servers, all of them have problem while using with patch fo ashift, sh we rollback path (for r255173) SH and all of them works for a week without that's problem's. Yesterday one of of servers was updated to stable/10 (beta1) SH SH wich include patch and after around 12 hours of works l2arc begin et errors like that SH SH kstat.zfs.misc.arcstats.l2_cksum_bad SH kstat.zfs.misc.arcstats.l2_io_error SH SH SH For now patch disabled in ower production. SH SH SH Please note we have very heavy load on zfs pool so 90GB arc and 3x180Gb L2arc have very big hit's on it on it. SH SH SH SSD used for cache's is intel ssd 530 series smart for all devices in in normal states's SH no bad values on it. SH SH Steven Hartland wrote: SH SH Have you confirmed the ashift changes are the actual cause of this SH SH by backing out just those changes and retesting on the same hardware. SH SH SH SH Also worth checking your disks smart values to confirm there are no SH SH visible signs of HW errors. SH SH SH SH Regards SH SH Steve SH SH SH SH - Original Message - SH SH From: Vitalij Satanivskij sa...@ukr.net SH SH To: Dmitriy Makarov suppor...@ukr.net SH SH Cc: Steven Hartland kill...@multiplay.co.uk; Justin T. Gibbs gi...@freebsd.org; Borja Marcos bor...@sarenet.es; SH SH freebsd-current@freebsd.org SH SH Sent: Wednesday, October 16, 2013 9:01 AM SH SH Subject: Re: ZFS secondarycache on SSD problem on r255173 SH SH SH SH SH SH Hello. SH SH SH SH Patch brocke cache functionality. SH SH SH SH Look at's Dmitriy's mail from Mon, 07 Oct 2013 21:09:06 +0300 SH SH SH SH With subject ZFS L2ARC - incorrect size and abnormal system load on r255173 SH SH SH SH As patch alredy in head and BETA it's not good. SH SH SH SH Yesterday we update one machine up to beta1 and forgot about patch. So 12 Hours and cache broken... :(( SH SH SH SH SH SH SH SH Dmitriy Makarov wrote: SH SH DM The attached patch by Steven Hartland fixes issue for me too. Thank you! SH SH DM SH SH DM SH SH DM --- Исходное сообщение --- SH SH DM От кого: Steven Hartland kill...@multiplay.co.uk SH SH DM Дата: 18 сентября 2013, 01:53:10 SH SH DM SH SH DM - Original Message - SH SH DM From: Justin T. Gibbs SH SH DM SH SH DM --- SH SH DM Дмитрий Макаров SH SH DM ___ SH SH DM freebsd-current@freebsd.org mailing list SH SH DM http://lists.freebsd.org/mailman/listinfo/freebsd-current SH SH DM To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org SH SH SH SH SH SH SH SH SH SH This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the SH event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any SH information contained in it. SH SH SH SH In the event of misdirection, illegible or incomplete transmission please telephone +44 845
Re: ZFS secondarycache on SSD problem on r255173
You'll have to be more specific. I don't have that email or know what list on which to search. Thanks, Justin On Oct 16, 2013, at 2:01 AM, Vitalij Satanivskij sa...@ukr.net wrote: Hello. Patch brocke cache functionality. Look at's Dmitriy's mail from Mon, 07 Oct 2013 21:09:06 +0300 With subject ZFS L2ARC - incorrect size and abnormal system load on r255173 As patch alredy in head and BETA it's not good. Yesterday we update one machine up to beta1 and forgot about patch. So 12 Hours and cache broken... :(( Dmitriy Makarov wrote: DM The attached patch by Steven Hartland fixes issue for me too. Thank you! DM DM DM --- Исходное сообщение --- DM От кого: Steven Hartland kill...@multiplay.co.uk DM Дата: 18 сентября 2013, 01:53:10 DM DM - Original Message - DM From: Justin T. Gibbs DM DM --- DM Дмитрий Макаров DM ___ DM freebsd-current@freebsd.org mailing list DM http://lists.freebsd.org/mailman/listinfo/freebsd-current DM To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
I took a quick look at arc.c and see that the trim_map_free() calls don't take into account ashift. I don't know if that has anything to do with your problem though. I would expect this would just make the trim less efficient, but I need to dig further. -- Justin On Oct 16, 2013, at 4:42 PM, Justin T. Gibbs gi...@freebsd.org wrote: You'll have to be more specific. I don't have that email or know what list on which to search. Thanks, Justin On Oct 16, 2013, at 2:01 AM, Vitalij Satanivskij sa...@ukr.net wrote: Hello. Patch brocke cache functionality. Look at's Dmitriy's mail from Mon, 07 Oct 2013 21:09:06 +0300 With subject ZFS L2ARC - incorrect size and abnormal system load on r255173 As patch alredy in head and BETA it's not good. Yesterday we update one machine up to beta1 and forgot about patch. So 12 Hours and cache broken... :(( Dmitriy Makarov wrote: DM The attached patch by Steven Hartland fixes issue for me too. Thank you! DM DM DM --- Исходное сообщение --- DM От кого: Steven Hartland kill...@multiplay.co.uk DM Дата: 18 сентября 2013, 01:53:10 DM DM - Original Message - DM From: Justin T. Gibbs DM DM --- DM Дмитрий Макаров DM ___ DM freebsd-current@freebsd.org mailing list DM http://lists.freebsd.org/mailman/listinfo/freebsd-current DM To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
- Original Message - From: Justin T. Gibbs gi...@freebsd.org I took a quick look at arc.c and see that the trim_map_free() calls don't take into account ashift. I don't know if that has anything to do with your problem though. I would expect this would just make the trim less efficient, but I need to dig further. This is actually done by TRIM_ZIO_END in trim_map_free. Regards Steve This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Ohh stupid question what hardware are you running this on, specifically what SSD's and what controller and if relavent what controller Firmware version? I wonder if you might have bad HW / FW, such as older LSI mps Firmware, which is know to causing corruption with some delete methods. Without the ashift fixes, you'll be sending short requests to the disk which will then ignore them, so I can see a potential corrilation there. You could rule this out by disabling ZFS TRIM with vfs.zfs.trim.enabled=0 in /boot/loader.conf Regards Steve This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
On Sep 17, 2013, at 4:53 PM, Steven Hartland kill...@multiplay.co.uk wrote: - Original Message - From: Justin T. Gibbs gi...@freebsd.org Sorry for being slow to chime in on this thread. I live in Boulder, CO and we've had a bit of rain. :-) Hope all is well your side, everyone safe and sound if may be little wetter than usual. I have been very fortunate. As Steven pointed out, the warning is benign, but does show that the code I committed to -current is not optimizing the allocation size for L2ARC devices. The fix for this is to find the best place during pool creation/load/import to call vdev_ashift_optimize() on L2ARC devices. I will look into this tomorrow, but feel free to suggest a good spot if you look at it before I can. The attached patch fixes this for me, not sure if its ideal place for it and consideration may be needed in combination with persistent l2arc. Persistent l2arc will require storing and checking the ashift recorded in the label so bugs and or quirk table changes don't confuse the system. But for now, with no persistent l2arc support in the tree, I think your placement is fine. I've submitted a request to RE so I can put your fix into head. -- Justin ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
- Original Message - From: Justin T. Gibbs gi...@freebsd.org Sorry for being slow to chime in on this thread. I live in Boulder, CO and we've had a bit of rain. :-) Hope all is well your side, everyone safe and sound if may be little wetter than usual. As Steven pointed out, the warning is benign, but does show that the code I committed to -current is not optimizing the allocation size for L2ARC devices. The fix for this is to find the best place during pool creation/load/import to call vdev_ashift_optimize() on L2ARC devices. I will look into this tomorrow, but feel free to suggest a good spot if you look at it before I can. The attached patch fixes this for me, not sure if its ideal place for it and consideration may be needed in combination with persistent l2arc. Regards Steve This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. zfs-l2arc-ashift.patch Description: Binary data ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
On Sep 13, 2013, at 2:18 PM, Steven Hartland wrote: This is a recent bit of code by Justin cc'ed, so he's likely the best person to investigate this one. Hmm. There is still a lot of confusion surrounding all this, and it's a time bomb waiting to explode. A friend had serious problems on 9.2 with the gnop trick in order to force a 4 KB block size. After a reboot, ZFS would insist on trying to find the .nop device for the ZIL which, of course, did no longer exist. In his case, for some reason, ZFS didn't identify the labelled gpt/name or gptpd/uuid devices as valid, and the pool wouldn't attach. Curiously, it did identify the L2ARC without problems. The cure was to disable GPT labels using loader.conf (I don't remember if he disabled kern.geom.label.gptid.enable, kern.geom.label.gpt.enable or both) in order to force it to use the classical daXpY nomenclature. As I said, this sounds like a time bomb in the works. There seems to be some confusion in the ZFS between the different naming schemes you can use for a disk partition right now ( daXpY, gpt label or gptid). Borja. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Cant say I've ever had a issue with gnop, but I haven't used it for some time. I did have a quick look over the weekend at your issue and it looks to me like warning for the cache is a false positive, as the vdev for cache doesn't report an ashift in zdb so could well be falling back to a default value. I couldn't reproduce the issue for log here, it just seems to work for me, can you confirm what ashift is reported for your devices using: zdb pool Regards Steve - Original Message - From: Borja Marcos bor...@sarenet.es To: Steven Hartland kill...@multiplay.co.uk Cc: Dmitryy Makarov suppor...@ukr.net; freebsd-current@freebsd.org; Justin T. Gibbs gi...@freebsd.org Sent: Monday, September 16, 2013 12:06 PM Subject: Re: ZFS secondarycache on SSD problem on r255173 On Sep 13, 2013, at 2:18 PM, Steven Hartland wrote: This is a recent bit of code by Justin cc'ed, so he's likely the best person to investigate this one. Hmm. There is still a lot of confusion surrounding all this, and it's a time bomb waiting to explode. A friend had serious problems on 9.2 with the gnop trick in order to force a 4 KB block size. After a reboot, ZFS would insist on trying to find the .nop device for the ZIL which, of course, did no longer exist. In his case, for some reason, ZFS didn't identify the labelled gpt/name or gptpd/uuid devices as valid, and the pool wouldn't attach. Curiously, it did identify the L2ARC without problems. The cure was to disable GPT labels using loader.conf (I don't remember if he disabled kern.geom.label.gptid.enable, kern.geom.label.gpt.enable or both) in order to force it to use the classical daXpY nomenclature. As I said, this sounds like a time bomb in the works. There seems to be some confusion in the ZFS between the different naming schemes you can use for a disk partition right now ( daXpY, gpt label or gptid). Borja. This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
Sorry for being slow to chime in on this thread. I live in Boulder, CO and we've had a bit of rain. :-) As Steven pointed out, the warning is benign, but does show that the code I committed to -current is not optimizing the allocation size for L2ARC devices. The fix for this is to find the best place during pool creation/load/import to call vdev_ashift_optimize() on L2ARC devices. I will look into this tomorrow, but feel free to suggest a good spot if you look at it before I can. -- Justin On Sep 16, 2013, at 6:43 AM, Steven Hartland kill...@multiplay.co.uk wrote: Thats correct the new code knows the what the ashift of the underlying disk should be so if it detects a vdev with a lower ashift you get this warning. I suspect the issue is that cache devices don't have an ashift, so when the check is done it gets a default value and hence the warning. I'd have to did some more to see how ashift in cache devices is or isn't implemented. Regards Steve - Original Message - From: Dmitriy Makarov suppor...@ukr.net To: Steven Hartland kill...@multiplay.co.uk Cc: Borja Marcos bor...@sarenet.es; freebsd-current@freebsd.org; Justin T. Gibbs gi...@freebsd.org Sent: Monday, September 16, 2013 1:29 PM Subject: Re[3]: ZFS secondarycache on SSD problem on r255173 And have to say that ashift of a main pool doesn't matter. I've tried to create pool with ashift 9 (default value) and with ashift 12 with creating gnops over gpart devices, export pool, destroy gnops, import pool. There is the same problem with cache device. There is no problem with ZIL devices, they reports ashift: 12 children[1]: type: 'disk' id: 1 guid: 6986664094649753344 path: '/dev/gpt/zil1' phys_path: '/dev/gpt/zil1' whole_disk: 1 metaslab_array: 67 metaslab_shift: 25 ashift: 12 asize: 4290248704 is_log: 1 create_txg: 22517 Problem with cache devices only, but in zdb output tere is nothing at all about them. --- Исходное сообщение --- От кого: Steven Hartland kill...@multiplay.co.uk Дата: 16 сентября 2013, 14:18:31 Cant say I've ever had a issue with gnop, but I haven't used it for some time. I did have a quick look over the weekend at your issue and it looks to me like warning for the cache is a false positive, as the vdev for cache doesn't report an ashift in zdb so could well be falling back to a default value. I couldn't reproduce the issue for log here, it just seems to work for me, can you confirm what ashift is reported for your devices using: zdb pool Regards Steve - Original Message - From: Borja Marcos --- Дмитрий Макаров --- Дмитрий Макаров This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: ZFS secondarycache on SSD problem on r255173
This is a recent bit of code by Justin cc'ed, so he's likely the best person to investigate this one. Regards Steve - Original Message - From: Dmitryy Makarov suppor...@ukr.net To: freebsd-current@freebsd.org Sent: Friday, September 13, 2013 12:16 PM Subject: ZFS secondarycache on SSD problem on r255173 Hello, FreeBSD Current. Have some trouble with adding SSD drive as secondarycache device on 10.0-CURRENT FreeBSD 10.0-CURRENT #1 r255173 : SSD INTEL SSDSC2BW180A4 dmesg: ada1 at ahcich2 bus 0 scbus4 target 0 lun 0 ada1: INTEL SSDSC2BW180A4 DC02 ATA-9 SATA 3.x device ada1: 300.000MB/s transfers (SATA 2.x, UDMA6, PIO 8192bytes) ada1: Command Queueing enabled ada1: 171705MB (351651888 512 byte sectors: 16H 63S/T 16383C) ada1: quirks=0x14K ada1: Previously was known as ad8 # diskinfo -v /dev/ada1 /dev/ada1 512 # sectorsize 180045766656 # mediasize in bytes (167G) 351651888 # mediasize in sectors 4096 # stripesize 0 # stripeoffset 348861 # Cylinders according to firmware. 16 # Heads according to firmware. 63 # Sectors according to firmware. BTDA326504N51802GN # Disk ident. Creating partition and adding: # gpart create -s gpt ada1 # gpart add -t freebsd-zfs -s3G -l zil0 ada1 # gpart add -t freebsd-zfs -l cache0 ada1 # zpool add disk1 log /dev/gpt/zil0 # zpool add disk1 cache /dev/gpt/cache0 After that have: log gpt/zil0 ONLINE 0 0 0 cache gpt/cache0 ONLINE 0 0 0 block size: 512B configured, 4096B native But why? ZFS can't properly set sector size for cache device (quirks=0x14K)? And have some strange behavior with ZIl partition on the same SSD - everything works! Okay, was trying to do some workaround with GNOP: # gnop create -S4096 /dev/gpt/cache0 # zpool add disk1 cache /dev/gpt/cache0.nop Seems to be OK cache gpt/cache0.nop ONLINE 0 0 0 But after reboot pool has status:One or more devices are configured to use a non-native block size. Expect reduced performance. cache gptid/39e40c33-1c4d-11e3-ac20-002590aa9548 ONLINE 0 0 0 block size: 512B configured, 4096B native And no gnop parts was found, and even gpart (in /dev/gpart) # gnop status # gnop: Command 'status' not available. # gpart show -l ada1 = 34 351651821 ada1 GPT (167G) 34 6 - free - (3.0k) 40 6291456 1 zil0 (3.0G) 6291496 345360352 2 cache0 (164G) 351651848 7 - free - (3.5k) # ls /dev/gpt/cache* zsh: no matches found: /dev/gpt/cache* Please, help to deal with this problem. Or what else information can be useful to find out this trouble. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org This e.mail is private and confidential between Multiplay (UK) Ltd. and the person or entity to whom it is addressed. In the event of misdirection, the recipient is prohibited from using, copying, printing or otherwise disseminating it or any information contained in it. In the event of misdirection, illegible or incomplete transmission please telephone +44 845 868 1337 or return the E.mail to postmas...@multiplay.co.uk. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org