Hi Dan,
That makes sense - the time between blacklist and magic comeback was
around 1 hour - thanks for the explanation. Is this is a safe default?
At eviction, the MDS takes all caps from the client away, so if it comes
back in an hour, doesn't it then write to files that it perhaps
For read-only workload, this should make no difference, since all read are from
SSD normally. But I think it’s still beneficial to writing, backfilling,
recovering. And also I will have some HDD only pools, so WAL/DB on SSD will
definitely improve performance for these pools. I will always put
在 2020年11月10日,02:26,Dave Hall 写道:
This thread caught my attention. I have a smaller cluster with a lot of OSDs
sharing the same SSD on each OSD node. I mentioned in an earlier post that I
found a statement in
We had some network problems (high packet drops) to some cephfs client
nodes that run ceph-fuse (14.2.13) against a Nautilus cluster (on
version 14.2.8). As a result a couple of clients got evicted (as one
would expect). What was really odd is that the clients were trying to
flush data they
Hi Mariusz, all
We have seen this issue as well, on redhat ceph 4 (I have an unresolved
case open). In our case, `radosgw-admin stat` is not a sufficient check to
guarantee that there are rados objects. You have to do a `rados stat` to
know that.
In your case, the object is ~48M in size, appears
Hi Andras,
The osd blocklist entries expire after 1hr by default:
Option("mon_osd_blacklist_default_expire", Option::TYPE_FLOAT,
Option::LEVEL_ADVANCED)
.set_default(1_hr)
.add_service("mon")
.set_description("Duration in seconds that blacklist entries for clients "
Hi,
Today while debugging something we had a few questions that might lead
to improving the cephfs forward scrub docs:
https://docs.ceph.com/en/latest/cephfs/scrub/
tldr:
1. Should we document which sorts of issues that the forward scrub is
able to fix?
2. Can we make it more visible (in docs)
Hi All,
Are there any methods for relocating an entire RGW bucket to a
different storage pool other than copying the contents?
Thanks,
Frank
___
ceph-users mailing list -- ceph-users@ceph.io
To unsubscribe send an email to ceph-users-le...@ceph.io
Problem Resolved: Reason - NoDangClue
I had the broken monitor sitting there trying to join and failing, just
watching the debug log scroll. I then stopped ceph-mon-01 and started it in
debug to watch the messages and also see if debug on ceph-mon-02 was able to
read it all.
Not only did it
Yes the OSDs are all bluestore. So does this mean that we can assign most
of the memory to the osd processes by setting the osd_memory_target?
>
> If your OSDs are all BlueStore, page cache isnât nearly as important as
> with Filestore.
>
>
>>
___
Hi,
Yeah the negative pid is interesting. AFAICT we use a negative pid to
indicate that the lock was taken on another host:
https://github.com/torvalds/linux/blob/master/fs/ceph/locks.c#L119
https://github.com/torvalds/linux/commit/9d5b86ac13c573795525ecac6ed2db39ab23e2a8
"Finally, we convert
Updating kernel versions has improved many Ceph-related things for me.
I don't use CentOS, but at a glance I see that you can get newer
kernels via "Elrepo". I would seriously consider doing so. Even
sticking with an LTS kernel, you can still get much newer kernels (up
to 5.4.75).
This thread caught my attention. I have a smaller cluster with a lot of
OSDs sharing the same SSD on each OSD node. I mentioned in an earlier post
that I found a statement in
https://docs.ceph.com/en/latest/rados/configuration/bluestore-config-ref/
indicating that if the SSD/NVMe in a node is
I feel lucky to have you on this one. ;-) Do you mean applying a
specific patch on 3.10 kernel? Or is this one too old to have it working
anyways.
Frédéric.
Le 09/11/2020 à 19:07, Luis Henriques a écrit :
Frédéric Nass writes:
Hi Luis,
Thanks for your help. Sorry I forgot about the
Luis,
I gave RHEL 8 and kernel 4.18 a try and it's working perfectly! \o/
Same commands, same mount options. Does anyone know why and if there's
any chances I can have this working with CentOS/RHEL 7 and 3.10 kernel?
Best regards,
Frédéric.
Le 09/11/2020 à 15:04, Frédéric Nass a écrit :
Frédéric Nass writes:
> Hi Luis,
>
> Thanks for your help. Sorry I forgot about the kernel details. This is latest
> RHEL 7.9.
>
> ~/ uname -r
> 3.10.0-1160.2.2.el7.x86_64
>
> ~/ grep CONFIG_TMPFS_XATTR /boot/config-3.10.0-1160.2.2.el7.x86_64
> CONFIG_TMPFS_XATTR=y
>
> upper directory /upperdir
My PGs are healthy now, but the underlying problem itself is not fixed. I was
interested if someone knew a fast fix to get the PGs complete right away.
The down OSDs have been shut down a long time ago and are sitting in a
different crush root. It was 1 OSD in an HDD pool that I'm re-organising
All,
I'm not sure if this is relevant here, but I recently tried to use
OverlayFS with an NFS share. It wouldn't work because NFS does not present
to the kernel as a block device. OverlayFS requires a block device
abstraction. If CephFS doesn't present as a block device you won't get it
to
I thought it might be related to reported issues where the MONs were
specified with IP:PORT, but that can be ruled out.
Does the current monmap match your actual setup? You wrote the keys
are correct, but maybe there's still a keyring left in 'ceph auth ls'?
Zitat von Paul Mezzanini :
It sounds like your client is able to reach the mon but not the OSD?
It needs to be able to reach all mons and all OSDs.
On Sun, Nov 8, 2020 at 4:29 AM Amudhan P wrote:
>
> Hi,
>
> I have mounted my cephfs (ceph octopus) thru kernel client in Debian.
> I get following error in "dmesg" when I try
Clients don't need the cluster IP because that's only for OSD <--> OSD
replication, no client traffic. But of course to be able to
communicate with Ceph the clients need a public IP, how else would
they contact the MON? Or did I misunderstand your setup?
Zitat von Amudhan P :
Hi,
I
Hi,
on my fresh deployed cephadm bootstrap node I can no longer run the ceph
command. It is just stuck:
[root@gedasvl02 ~]# ceph orch dev ls
INFO:cephadm:Inferring fsid c7879f24-1f90-11eb-8ba2-005056b703af
INFO:cephadm:Using recent ceph image docker.io/ceph/ceph:v15
[root@gedasvl02 ~]#
Thank you, Istvan, and Amit, who also replied.
I had tried three Ceph versions, including (after your suggestion), the
latest, but in my case it wasn't that.
What the problem was, I don't know, but I changed from using VMs to test
this functionality last week to using clusters of real hardware
Hi,
Couple of questions came up which is not really documented anywhere, hopefully
someone knows the answers:
1.
Is there a way to see the replication queue? I want to create metrics like is
there any delay in the replication etc ...
2.
Is the replication FIFO?
3.
Actually how a replication
Hi Luis,
Thanks for your help. Sorry I forgot about the kernel details. This is
latest RHEL 7.9.
~/ uname -r
3.10.0-1160.2.2.el7.x86_64
~/ grep CONFIG_TMPFS_XATTR /boot/config-3.10.0-1160.2.2.el7.x86_64
CONFIG_TMPFS_XATTR=y
upper directory /upperdir is using xattrs
~/ ls -l
We are having the exact same problem (also Octopus). The object is
listed by s3cmd, but trying to download it results in a 404 error.
radosgw-admin object stat shows that the object still exists. Any
further ideas how I can restore access to this object?
(Sorry if this is a duplicate, but it
Hi all,
MDS version v14.2.11
Client kernel 3.10.0-1127.19.1.el7.x86_64
We are seeing a strange issue with a dovecot use-case on cephfs.
Occasionally we have dovecot reporting a file locked, such as:
Nov 09 13:55:00 dovecot-backend-00.cern.ch dovecot[27710]:
imap(reguero)<23945>: Error: Mailbox
Hi Anthony
>
> Did you add a bunch of data since then, or change the Ceph release? Do
> you have bluefs_buffered_io set to false?
>
>
We did not change ceph release in the meantime. It is very welll possible
that the delays were just not noticed during out previous maintenances.
Frédéric Nass writes:
> Hello,
>
> I would like to use a cephfs snapshot as a read/write volume without having to
> clone it first as the cloning operation is - if I'm not mistaken - still
> inefficient as of now. This is for a data restore use case with Moodle
> application needing a writable
Hello,
I would like to use a cephfs snapshot as a read/write volume without
having to clone it first as the cloning operation is - if I'm not
mistaken - still inefficient as of now. This is for a data restore use
case with Moodle application needing a writable data directory to start.
The
30 matches
Mail list logo