Re: Interactive/Watch ceph -s/-w/status

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 




On 06/14/2018 07:05 PM, John Spray wrote:
> On Wed, Jun 6, 2018 at 2:29 PM, Wido den Hollander <wido@xxxxxxxx> wrote:
>> Hi,
>>
>> Previous to Luminous you could run 'ceph -w' and get a continues output
>> of how the cluster is performing. You would get IOps and such.
>>
>> To get this done with Luminous you need to do something like:
>>
>> $ watch ceph -s
>>
>> Although this works it is less efficient then -w used to be:
>>
>> 1. Spawn Python interpreter
>> 2. Create connection with the MON(s)
>> 3. Fetch OSDMap
>> 4. Fetch statistics
>>
>> And do this all every 2 seconds (or which interval you set).
>>
>> So I was thinking about this:
>>
>> $ ceph -s <interval>
>>
>> If you provide a interval the process will not exit but it will refresh
>> the screen every X seconds while keeping the Python interpreter live and
>> also keep the MON connection imitated.
>>
>> It's actually just running the mon command 'status' every X seconds and
>> refreshing the screen.
>>
>> Although we have the 'iostat' Mgr module in Mimic it doesn't do exactly
>> the same as status does.
>>
>> I think it's the status Mgr Module which is responsible for this, isn't it?
> 
> The status module is something else (the "fs status" and "osd status"
> commands, and any other handy things anyone wanted to add).
> 
> If one literally just wants a live-refreshed view of "ceph -s", then
> "watch ceph -s" is IMO quite reasonable, and not inefficient enough to
> justify creating a built in version (unless we're feeling the pain of
> the continuous reconnections in a way that I haven't thought of?)
> 
> However, I could see the need for some variations on the theme of the
> iostat command, that had more of the health and status info
> (especially pg states) rather than performance-focussed output.  I
> find the line-streaming output more handy than a full screen refresh,
> because I have a few lines of history on the screen.
> 

Yes, that's why I liked the old 'ceph -w' as I would keep my updated on
how the cluster is performing. Instant PG updates, something you really
want to know while you are in the middle of a situation.

Can Mgr modules stream output instead of refreshing?

Wido

> John
> 
>>
>> Wido
>> --
>> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
>> the body of a message to majordomo@xxxxxxxxxxxxxxx
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html



[Index of Archives]     [CEPH Users]     [Ceph Large]     [Information on CEPH]     [Linux BTRFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux