mons")
HEALTH_OK, tunables optimal.
What is it?
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Dzianis Kahanovich пишет:
After upgrading 0.80.1 to 0.80.3 I see many regular messages on every OSD log:
2014-07-15 19:44:48.292839 7fa5a659f700 0 osd.5 62377 crush map has features
2199057072128, adjusting msgr requires for mons
(constant part: "crush map has features 21990570
Dzianis Kahanovich пишет:
Dzianis Kahanovich пишет:
After upgrading 0.80.1 to 0.80.3 I see many regular messages on every OSD log:
2014-07-15 19:44:48.292839 7fa5a659f700 0 osd.5 62377 crush map has features
2199057072128, adjusting msgr requires for mons
(constant part: "crush ma
I plan to migrate cluster from straw to straw2 mapping. Ceph and kernels is up
to date (kernel 4.1.0), so I want to change directly in crush map srew to straw2
and load changed crush map (by steps - per host and rack). Are this relative
safe and must be remapped runtime?
--
WBR, Dzianis
Wido den Hollander пишет:
On 06/24/2015 05:11 PM, Dzianis Kahanovich wrote:
I plan to migrate cluster from straw to straw2 mapping. Ceph and kernels
is up to date (kernel 4.1.0), so I want to change directly in crush map
srew to straw2 and load changed crush map (by steps - per host and
rack
ctly prohibited.
If you have received this communication in error, please notify the sender by
telephone or e-mail (as shown above) immediately and destroy any and all copies
of this message in your possession (whether hard copies or electronically
stored copies).
_
UxKRK9O2hNSzFf6+1heT
yQ3L
=DreJ
-END PGP SIGNATURE-
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.
Reading on SourceForge blog, there are experienced ceph corruption. IMHO there
will be good idea to know technical details. Version, what happened...
http://sourceforge.net/blog/sourceforge-infrastructure-and-service-restoration/
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http
51
> 851
> 851
> 1451
> 362
> 851
> 851
> 851
> ...
>
> The problem does not seem to be related to Samba. If I copy the files to an
> XFS volume and export that, things look fine.
>
> Thanks
> Joerg Henne
>
> __
PS I start to use this patches with samba 4.1. IMHO some of problems may (or
must) be solved not inside vfs code, but outside - in samba kernel, but I still
use both in samba 4.2.3 without verification.
Dzianis Kahanovich пишет:
> I use cephfs over samba vfs and have some issues.
>
> 1)
st in Oracle linux or SuSE, but I use Gentoo and simple
heartbeat, no corosync, so I don't want to work too much).
So, if you need no byte-range locking, I suggest to use OCFS2 with simple O2CB
stack.
--
WBR, Dzianis Kahanovich AKA Denis Kag
serious overheads?
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
t of the stack now.
Are "namespace" can be used for user quota for cephfs (not usual "quota" with
capabilities, but just per-user or per-dir space limit)?
I remember thread about directory quotas branch, but find nothing related in
current tree. Are this project activ
ing, but quota is required. Doing this with pools enabling for
users other ceph features, but there are just perspective goal, not for current
(our) users.
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-use
and mount this point via fuse.
PS For ceph developers: trying this for qouta (with "ceph osd pool set-quota")
semi-working: on quota overflow - nothing limited, but "ceph health" show
warning. In case of no other ways to quota, it may qualified as "bug" and not
too ac
> performance drastically.
I see some sense in RAID 0: single ceph-osd daemon per node (but still
disk-per-osd self). But if you have relative few [planned] cores per task on
node - you can think about it.
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
ut it.
>
> Raid-0: single disk failure kills the entire filesystem, off-lines the osd and
> triggers a cluster-wide resync. Actual raid: single disk failure does not
> affect
> the cluster in any way.
Usually data distributed per-host, so whole array f
cache. All other jobs (striping, fault tolerance) are
Ceph's. Better to buy many biggest possible disks and insert it into many usual
SATA machines.
And usually I kill hardware RAID on new machines and start mdadm (if there are
single-node Linux server) - to avoid painful games with various hardware.
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
etter read/write
> balance if the osd number is increased.
>
> But seems the pg number cannot be changed on the fly. It's fixed when the
> pool is created. Am I right?
>
>
>
>
>
>
> _____
ause perfomance
degradation) or there will be mostly one thread?
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
How to avoid "slow requests" on rbd v1 snapsot delete? Time ago it looks solved,
but on "emperor" seen again.
Are migrating to rbd v2 can solve it?
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
__
mds.b]
host = megaserver3
[mds.c]
host = megaserver4
(I trying to unswitch all non-defaults, IMHO no results - fixme)
Or may be I need special care on mds stop (now - SIGKILL).
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unib
John Spray пишет:
On Tue, Oct 6, 2015 at 11:43 AM, Dzianis Kahanovich
wrote:
Short: how to sure avoid (if possible) fs freezes on 1 of 3 mds rejoin?
ceph version 0.94.3-242-g79385a8 (79385a85beea9bccd82c99b6bda653f0224c4fcd)
I moving 2 VM clients from ocfs2 (starting to deadlock VM on
PS This is standard 3 node (MON+MDS+OSDs - initial 3x setup) cluster + 1 OSDs
later node. Nothing special. OSDs balanced near equal size per host.
Dzianis Kahanovich пишет:
John Spray пишет:
On Tue, Oct 6, 2015 at 11:43 AM, Dzianis Kahanovich
wrote:
Short: how to sure avoid (if possible) fs
mds op history size = 50
# vs. laggy beacon
mds decay halflife = 9
mds beacon interval = 8
mds beacon grace = 30
[mds.a]
host = megaserver1
[mds.b]
host = megaserver3
[mds.c]
host = megaserver4
(I trying to unswitch all non-default
John Spray пишет:
On Tue, Oct 6, 2015 at 1:22 PM, Dzianis Kahanovich
wrote:
Even now I remove "mds standby replay = true":
e7151: 1/1/1 up {0=b=up:active}, 2 up:standby
Cluster stuck on KILL active mds.b. How to correctly stop mds to get
behaviour like on MONs - leader->down
Sorry, skipped some...
John Spray пишет:
On Tue, Oct 6, 2015 at 1:22 PM, Dzianis Kahanovich
wrote:
Even now I remove "mds standby replay = true":
e7151: 1/1/1 up {0=b=up:active}, 2 up:standby
Cluster stuck on KILL active mds.b. How to correctly stop mds to get
behaviour like on MON
John Spray пишет:
On Tue, Oct 6, 2015 at 2:21 PM, Dzianis Kahanovich
wrote:
John Spray пишет:
On Tue, Oct 6, 2015 at 1:22 PM, Dzianis Kahanovich
wrote:
Even now I remove "mds standby replay = true":
e7151: 1/1/1 up {0=b=up:active}, 2 up:standby
Cluster stuck on KILL active mds
0700 7 mds.0.cache.dir(19e3a8c) already
fetching; waiting
2015-10-06 23:43:40.929537 7f255eb50700 7 mds.0.cache.dir(19a66a4) already
fetching; waiting
2015-10-06 23:43:40.936432 7f255eb50700 7 mds.0.cache.dir(19c8188) already
fetching; waiting
2015-10-06 23:43:40.975802 7f255ca4b700 -1
m
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
2 additional out-of-cluster (service) nodes:
4.1.8 (now 4.2.3) kernel mount;
4.1.0 both mounts;
3) 2 VMs:
kernel mounts (most active: web & mail);
4.2.3;
fuse mounts - same version with ceph;
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibe
);
looks like freezed by CLONE_VFORK (B) freeze (A) & others (B), (but sometime on
PREEMPT, always - on PREEMPT_NONE).
I will restart mds this night, will look to restart time.
Dzianis Kahanovich пишет:
Yan, Zheng пишет:
It seems you have 16 mounts. Are you using kernel client or fuse
cli
on apache root. In this place was (no more now) described
CLONE_FS -> CLONE_VFORK deadlocks. But 4.2.3 installed just before tests, was
4.1.8 with similar effects (but log from 4.2.3 on VM clients).
Waith this night for MDSs restart.
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, ht
GBZSInFgUpPwUQKvF5OSGb3YQlzofUy5Es+wH3ccqZ+mlIY
szuBLAtN6zhFFPCs6016hiragiUhLk97PItXaKdDJKecuyRdShlJrXJmtX+j
NdM14TkBPTiLtAd/IZEEhIIpdvQH8YSl3LnEZ5gywggaY4Pk3JLFIJPgLpEb
T8hJnuiaQaYxERQ0nRoBL4LAXARseSrOuVt2EAD50Yb/5JEwB9FQlN758rb1
AE/xhpK6d53+RlkPODKxXx816hXvDP6NADaC78XGm
Dzianis Kahanovich пишет:
Waith this night for MDSs restart.
Now (3 active mds, balanced, nodes upgrade from git & restart - first - mon,
next - mds+osd, without debug) - reduced time, but anymore clients wait to
rejoin complete of related mds. May be need rebalance with maximum &
les.
during mds failover, mds needs to open these files, which take a long
time.
Can some kind of cache improve behaviour?
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
h
h-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.c
ng) and pg
verification after restart, at least to avoid "active" for broken pg.
PS 0.94.5
PPS 4.3.0 not stuck on mount, but xfs_repair still required.
PPPS Use swap and avoid forced kill.
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
aged
stopped
data_pools 5
metadata_pool 6
inline_data disabled
3104110:10.227.227.103:6800/14627 'a' mds.0.5436 up:active seq 30
3084126:10.227.227.104:6800/24069 'c' mds.0.0 up:standby-replay seq 1
If standby-replay false - all OK: 1/1/1 up {0=a=up
Gregory Farnum пишет:
> On Thu, Jun 30, 2016 at 1:03 PM, Dzianis Kahanovich wrote:
>> Upgraded infernalis->jewel (git, Gentoo). Upgrade passed over global
>> stop/restart everything oneshot.
>>
>> Infernalis: e5165: 1/1/1 up {0=c=up:active}, 1 up:standby-replay, 1
nning mds) .
>
> This is the first time you've upgraded your pool to jewel right?
> Straight from 9.X to 10.2.2?
>
Yes
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.by/
___
ceph-users mailing list
ceph-users@lis
anually.
If you say cmake is preferred by developers (and can solve some of this) - I
will try rework Gentoo ebuild to it (my own and will report into Gentoo
bugzilla).
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.by/
___
ceph-use
al") inconsistent PGs to "size" in write task (if
code enables it).
So write requests will be always protected from data loss (of course, still
possibility to invert written and offline OSDs in one pass in large cluster, but
this is minimal care for mind about min_size).
--
WBR,
I have 4 scrub errors (3 PGs - "found clone without head"), on one OSD. Not
repairing. How to repair it exclude re-creating of OSD?
Now it "easy" to clean+create OSD, but in theory - in case there are multiple
OSDs - it may cause data lost.
--
WBR, Dzianis Kahanovich AKA De
Dzianis Kahanovich пишет:
> I have 4 scrub errors (3 PGs - "found clone without head"), on one OSD. Not
> repairing. How to repair it exclude re-creating of OSD?
>
> Now it "easy" to clean+create OSD, but in theory - in case there are multiple
> OSDs - it m
Does somebody try ceph + suspend|hibernate (for UPS power-off)? Can it cause
problems with ceph sync in case async poweroff? Fear to try on production (v2)
first!
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph
://www.mermaidconsulting.com/
>> ___
>> ceph-users mailing list
>> ceph-users@lists.ceph.com
>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
> ___
> ceph-users mailing list
> cep
Freezer in guest agent? Exists or todo?
Dzianis Kahanovich пишет:
> IMHO interaction QEMU & kernel's FREEZER (part of hibernation & cgroups) can
> solve many of problems. It can be done via QEMU host-2-guest sockets and
> scripts
> or embedded into virtual hardwar
heard of such
> mechanism except academic projects.
>
> On Sat, May 18, 2013 at 5:48 PM, Dzianis Kahanovich
> wrote:
>> IMHO interaction QEMU & kernel's FREEZER (part of hibernation & cgroups) can
>> solve many of problems. It can be done via QEMU host-2-gue
gt;> IMHO interaction QEMU & kernel's FREEZER (part of hibernation & cgroups) can
>> solve many of problems. It can be done via QEMU host-2-guest sockets and
>> scripts
>
> That would require a cooperating VM. What I was looking at was how to do this
> for non-coopera
=1).
But try to do nothing first and try to install latest version. And keep your
vote to issue #4937 to force developers.
Olivier Bonvalet пишет:
> Le mardi 07 mai 2013 à 15:51 +0300, Dzianis Kahanovich a écrit :
>> I have 4 scrub errors (3 PGs - "found clone without head"
Olivier Bonvalet пишет:
>
> Le lundi 20 mai 2013 à 00:06 +0200, Olivier Bonvalet a écrit :
>> Le mardi 07 mai 2013 à 15:51 +0300, Dzianis Kahanovich a écrit :
>>> I have 4 scrub errors (3 PGs - "found clone without head"), on one OSD. Not
>>> repairing. How
gt;
> (I'm hopeful because virtio-scsi had multi-queue support for a while,
> and someone reported increased IOPS even with RBD devices behind those.)
>
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
.103:6802/5237 16 : cluster [ERR]
repair 3.4e 3:73d0516f:::rbd_data.2d2082ae8944a.3239:2368 is an
unexpected clone
2016-09-09 17:24:26.490788 osd.1 10.227.227.103:6802/5237 17 : cluster [ERR]
3.4e repair 0 missing, 1 inconsistent objects
2016-09-09 17:24:26.490807 osd.1 10.227.22
Dzianis Kahanovich пишет:
>
> I have 1 active+clean+inconsistent PG (from metadata pool) without real error
> reporting and any other symphtoms. All 3 copies same (md5sum). Deep-scrub,
> repair, etc just say "1 errors 0 fixed" in the end. I remember, it PG may be
> han
t?
PS "setuser match path = /var/lib/ceph/$type/$cluster-$id" added to config.
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
I have content for apache 2.4 in cephfs, trying to be scalable, "EnableMMAP On".
Some environments known as not friendly for MMAP for SMP scalability (more
locks). What cephfs-specific recommendations about apache's EnableMMAP setting?
--
WBR, Dzianis Kahanovich AKA Denis K
g now after repair too).
PS hammer from git.
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Fix: affected not only Megaraid SAS. For tested time:
Affected: MegaRAID SAS 2108, Intel 82801JI
Unaffected: Intel C602
Both Intel's in AHCI mode.
So, hardware possible not important.
Dzianis Kahanovich пишет:
> This issue was fixed by "xfs_repair -L".
>
> 1) Megaraid
eph/commit/24de350d936e5ed70835d0ab2ad6b0b4f506123f.patch
, previous incident was older & without patch.
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
PS Now I stop this mds, active migrated and warning removed. Cannot try more.
Dzianis Kahanovich пишет:
> John Spray пишет:
>
>>> Looks happened both time at night - probably on long backup/write operations
>>> (something like compressed local root backup to cephfs). Al
e servers with cluster) I mounts with
"wsize=131072,rsize=131072,write_congestion_kb=128,readdir_max_bytes=131072"
(and net.ipv4.tcp_notsent_lowat = 131072) to conserve RAM. Obtaining good
servers for VMs I remove it. May be better turn it back for better congestion
quantum.
--
W
l distros
reality...
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich, http://mahatma.bspu.unibel.by/
___
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
threads numbers. Scheduler=noop. size=3 min_size=2
No same problem with fuse.
Looks like broken or unbalanced congestion mechanism or I don't know how to
moderate it. write_congestion_kb trying low (=1) - nothing interesting.
--
WBR, Dzianis Kahanovich AKA Denis Kaganovich,
hus creating
> very large writes (how much memory does your test machine have) when it
> gets flushed.
I bound all read/write values in kernel client more then fuse.
Mostly I understand - problem are fast write & slow HDDs. But IMHO some
mechanisms must prevent it (congestion-like). And
Dzianis Kahanovich пишет:
> Christian Balzer пишет:
>
>>> New problem (unsure, but probably not observed in Hammer, but sure in
>>> Infernalis): copying large (tens g) files into kernel cephfs (from
>>> outside of cluster, iron - non-VM, preempt kernel) - make
66 matches
Mail list logo
