Hi Tarek, Am 29.05.19 um 18:49 schrieb Tarek Zegar: > Hi Oliver, > > Thank you for the response, I did ensure that min-client-compact-level is > indeed Luminous (see below). I have no kernel mapped rbd clients. Ceph > versions reports mimic. Also below is the output of ceph balancer status. One > thing to note, I did enable the balancer after I already filled the cluster, > not from the onset. I had hoped that it wouldn't matter, though your comment > "if the compat-level is too old for upmap, you'll only find a small warning > about that in the logfiles" leaves me to believe that it will *not* work in > doing it this way, please confirm and let me know what message to look for in > /var/log/ceph.
it should also work well on existing clusters - we have also used it on a
Luminous cluster after it was already half-filled, and it worked well - that's
what it was made for ;-).
The only issue we encountered was that the client-compat-level needed to be set
to Luminous before enabling the balancer plugin, but since you can always
disable and re-enable a plugin,
this is not a "blocker".
Do you see anything in the logs of the active mgr when disabling and
re-enabling the balancer plugin?
That's how we initially found the message that we needed to raise the
client-compat-level.
Cheers,
Oliver
>
> Thank you!
>
> root@hostadmin:~# ceph balancer status
> {
> "active": true,
> "plans": [],
> "mode": "upmap"
> }
>
>
>
> root@hostadmin:~# ceph features
> {
> "mon": [
> {
> "features": "0x3ffddff8ffacfffb",
> "release": "luminous",
> "num": 3
> }
> ],
> "osd": [
> {
> "features": "0x3ffddff8ffacfffb",
> "release": "luminous",
> "num": 7
> }
> ],
> "client": [
> {
> "features": "0x3ffddff8ffacfffb",
> "release": "luminous",
> "num": 1
> }
> ],
> "mgr": [
> {
> "features": "0x3ffddff8ffacfffb",
> "release": "luminous",
> "num": 3
> }
> ]
> }
>
>
>
>
> Inactive hide details for Oliver Freyermuth ---05/29/2019 11:13:51 AM---Hi
> Tarek, what's the output of "ceph balancer status"?Oliver Freyermuth
> ---05/29/2019 11:13:51 AM---Hi Tarek, what's the output of "ceph balancer
> status"?
>
> From: Oliver Freyermuth <[email protected]>
> To: [email protected]
> Date: 05/29/2019 11:13 AM
> Subject: [EXTERNAL] Re: [ceph-users] Balancer: uneven OSDs
> Sent by: "ceph-users" <[email protected]>
>
> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
>
>
> Hi Tarek,
>
> what's the output of "ceph balancer status"?
> In case you are using "upmap" mode, you must make sure to have a
> min-client-compat-level of at least Luminous:
> http://docs.ceph.com/docs/mimic/rados/operations/upmap/
> Of course, please be aware that your clients must be recent enough
> (especially for kernel clients).
>
> Sadly, if the compat-level is too old for upmap, you'll only find a small
> warning about that in the logfiles,
> but no error on terminal when activating the balancer or any other kind of
> erroneous / health condition.
>
> Cheers,
> Oliver
>
> Am 29.05.19 um 17:52 schrieb Tarek Zegar:
>> Can anyone help with this? Why can't I optimize this cluster, the pg counts
>> and data distribution is way off.
>> __________________
>>
>> I enabled the balancer plugin and even tried to manually invoke it but it
>> won't allow any changes. Looking at ceph osd df, it's not even at all.
>> Thoughts?
>>
>> root@hostadmin:~# ceph osd df
>> ID CLASS WEIGHT REWEIGHT SIZE USE AVAIL %USE VAR PGS
>> 1 hdd 0.00980 0 0 B 0 B 0 B 0 0 0
>> 3 hdd 0.00980 1.00000 10 GiB 8.3 GiB 1.7 GiB 82.83 1.14 156
>> 6 hdd 0.00980 1.00000 10 GiB 8.4 GiB 1.6 GiB 83.77 1.15 144
>> 0 hdd 0.00980 0 0 B 0 B 0 B 0 0 0
>> 5 hdd 0.00980 1.00000 10 GiB 9.0 GiB 1021 MiB 90.03 1.23 159
>> 7 hdd 0.00980 1.00000 10 GiB 7.7 GiB 2.3 GiB 76.57 1.05 141
>> 2 hdd 0.00980 1.00000 10 GiB 5.5 GiB 4.5 GiB 55.42 0.76 90
>> 4 hdd 0.00980 1.00000 10 GiB 5.9 GiB 4.1 GiB 58.78 0.81 99
>> 8 hdd 0.00980 1.00000 10 GiB 6.3 GiB 3.7 GiB 63.12 0.87 111
>> TOTAL 90 GiB 53 GiB 37 GiB 72.93
>> MIN/MAX VAR: 0.76/1.23 STDDEV: 12.67
>>
>>
>> root@hostadmin:~# osdmaptool om --upmap out.txt --upmap-pool rbd
>> osdmaptool: osdmap file 'om'
>> writing upmap command output to: out.txt
>> checking for upmap cleanups
>> upmap, max-count 100, max*deviation 0.01 <---really? It's not even close to
>> 1% across the drives*
>> limiting to pools rbd (1)
>> *no upmaps proposed*
>>
>>
>> ceph balancer optimize myplan
>> Error EALREADY: Unable to find further optimization,or distribution is
>> already perfect
>>
>>
>> _______________________________________________
>> ceph-users mailing list
>> [email protected]
>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>>
>
> /(See attached file:
> smime.p7s)/_______________________________________________
> ceph-users mailing list
> [email protected]
> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>
>
>
smime.p7s
Description: S/MIME Cryptographic Signature
_______________________________________________ ceph-users mailing list [email protected] http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
