Re: Partition loss event

[Ilya Kasnacheev]

Dear fellows!

I think we have a problem: when events were introduced, we were talking
about high-bandwdith events which may overflow your nodes if you
accidentally turn them on.

However, now we have a bunch of low-bandwidth events, such as:
EVT_CHECKPOINT_SAVED
EVT_CHECKPOINT_LOADED
EVT_CHECKPOINT_REMOVED
EVT_NODE_JOINED
EVT_NODE_LEFT
EVT_NODE_FAILED
EVT_NODE_SEGMENTED
EVT_CACHE_REBALANCE_STARTED
EVT_CACHE_REBALANCE_STOPPED
EVT_CACHE_REBALANCE_PART_LOADED
EVT_CACHE_REBALANCE_PART_UNLOADED
EVT_CACHE_REBALANCE_OBJECT_LOADED
EVT_CACHE_REBALANCE_OBJECT_UNLOADED
EVT_CACHE_REBALANCE_PART_DATA_LOST
EVT_CACHE_REBALANCE_PART_SUPPLIED
EVT_CACHE_REBALANCE_PART_MISSED
EVT_CLIENT_NODE_DISCONNECTED
EVT_CLIENT_NODE_RECONNECTED
EVT_WAL_SEGMENT_ARCHIVED
EVT_WAL_SEGMENT_COMPACTED
EVT_CLUSTER_ACTIVATED
EVT_CLUSTER_DEACTIVATED
EVT_PAGE_REPLACEMENT_STARTED

I suggest we enable these events by default, since I fail to see how this
may ever cause problems, but it will definitely decrease confusion
surrounding events.

WDYT?

Regards,
--
Ilya Kasnacheev


чт, 1 авг. 2019 г. в 15:18, balazspeterfi <[hidden email]>:

> Hi Alexandr,
>
> Thanks, that was the missing part. It would be nice to mention it in the
> docs I guess as it's quite easy to miss it.
>
> Regards,
> Balazs
>
>
>
> --
> Sent from: http://apache-ignite-users.70518.x6.nabble.com/
>

[Ivan Pavlukhin]

Ilya,

I am not sure that enabling subset of events will fit all needs. If we
would like to do so then it might be good idea to make it clear in API
that certain events are enabled by default (e.g. put them into
separate class).

Also we should be careful with backward compatibility, some kind of
events look more like debug stuff, e.g.
EVT_CACHE_REBALANCE_OBJECT_LOADED. Is it relevant for historical
rebalance? What does it mean for rebalancing whole partition files?
How user is going to use it?

And a care stil should be put to performance, aforementioned
EVT_CACHE_REBALANCE_OBJECT_LOADED can introduce non-negliable impact,
cannot it?

Also, we can employ some soft means like printing a warning if a
listener is registered for disabled event.

And a last point about code and javadocs. There is a line "Note that
certain events are required for Ignite's internal operations and such
events will still be generated". Perhaps we can provide a list of such
events in docs or do not expose them to a user listeners. And a bit of
mess. IgniteConfiguration#getIncludeEventTypes claims "Note that by
default all events in Ignite are disabled". While EventType says "Note
that by default all events in Ignite are enabled and therefore
generated and stored by whatever event storage SPI is configured".

чт, 1 авг. 2019 г. в 17:42, Ilya Kasnacheev <[hidden email]>:

>
> Dear fellows!
>
> I think we have a problem: when events were introduced, we were talking
> about high-bandwdith events which may overflow your nodes if you
> accidentally turn them on.
>
> However, now we have a bunch of low-bandwidth events, such as:
> EVT_CHECKPOINT_SAVED
> EVT_CHECKPOINT_LOADED
> EVT_CHECKPOINT_REMOVED
> EVT_NODE_JOINED
> EVT_NODE_LEFT
> EVT_NODE_FAILED
> EVT_NODE_SEGMENTED
> EVT_CACHE_REBALANCE_STARTED
> EVT_CACHE_REBALANCE_STOPPED
> EVT_CACHE_REBALANCE_PART_LOADED
> EVT_CACHE_REBALANCE_PART_UNLOADED
> EVT_CACHE_REBALANCE_OBJECT_LOADED
> EVT_CACHE_REBALANCE_OBJECT_UNLOADED
> EVT_CACHE_REBALANCE_PART_DATA_LOST
> EVT_CACHE_REBALANCE_PART_SUPPLIED
> EVT_CACHE_REBALANCE_PART_MISSED
> EVT_CLIENT_NODE_DISCONNECTED
> EVT_CLIENT_NODE_RECONNECTED
> EVT_WAL_SEGMENT_ARCHIVED
> EVT_WAL_SEGMENT_COMPACTED
> EVT_CLUSTER_ACTIVATED
> EVT_CLUSTER_DEACTIVATED
> EVT_PAGE_REPLACEMENT_STARTED
>
> I suggest we enable these events by default, since I fail to see how this
> may ever cause problems, but it will definitely decrease confusion
> surrounding events.
>
> WDYT?
>
> Regards,
> --
> Ilya Kasnacheev
>
>
> чт, 1 авг. 2019 г. в 15:18, balazspeterfi <[hidden email]>:
>
> > Hi Alexandr,
> >
> > Thanks, that was the missing part. It would be nice to mention it in the
> > docs I guess as it's quite easy to miss it.
> >
> > Regards,
> > Balazs
> >
> >
> >
> > --
> > Sent from: http://apache-ignite-users.70518.x6.nabble.com/
> >

--
Best regards,
Ivan Pavlukhin