[tahoe-lafs-trac-stream] [Tahoe-LAFS] #486: add munin graphs of server-operation latency

[Tahoe-LAFS]

#486: add munin graphs of server-operation latency
------------------------------+------------------------------------
     Reporter:  warner        |      Owner:
         Type:  enhancement   |     Status:  closed
     Priority:  major         |  Milestone:  undecided
    Component:  code-storage  |    Version:  1.1.0
   Resolution:  wontfix       |   Keywords:  performance statistics
Launchpad Bug:                |
------------------------------+------------------------------------
Changes (by exarkun):

 * status:  new => closed
 * resolution:   => wontfix


Old description:

> We recently added code to collect server-side latency stats (how long it
> takes from the beginning of remote_get_buckets to just before it returns,
> etc), and to make these available to the stats-gatherer.
>
> We should also write some munin plugins to graph them over time. This may
> be useful to see how the server's disk contention is changing over time:
> with a certain level of traffic, does it take longer to perform a single
> read operation? Do different local filesystem types (ext3, reiserfs, zfs)
> run faster than others?

New description:

 We recently added code to collect server-side latency stats (how long it
 takes from the beginning of remote_get_buckets to just before it returns,
 etc), and to make these available to the stats-gatherer.

 We should also write some munin plugins to graph them over time. This may
 be useful to see how the server's disk contention is changing over time:
 with a certain level of traffic, does it take longer to perform a single
 read operation? Do different local filesystem types (ext3, reiserfs, zfs)
 run faster than others?

--

Comment:

 This ticket is about operational visibility for operations that are no
 longer operational.

--
Ticket URL: <https://tahoe-lafs.org/trac/tahoe-lafs/ticket/486#comment:3>
Tahoe-LAFS <https://Tahoe-LAFS.org>
secure decentralized storage