Hello Bruno / Nick, Thanks for the great feedback (again)! I want to give my thoughts to them and reply together here since some of them are correlated.
1) This KIP is indeed based on the separated restoration thread work, trying to give more visibility as we move restoration to dedicated threads. At the moment the number of restoration threads is not configurable, and is correlated to the num.main threads, but in the future we would make the num.restore.threads either configurable or at least not correlated to num.main threads. So from that perspective it's indeed not very reasonable to have those metrics on the thread-level since Streams will try to abstract away from users. However, from our initial framework on Streams metrics ( https://cwiki.apache.org/confluence/display/KAFKA/KIP-444%3A+Augment+metrics+for+Kafka+Streams), task-level metrics would by default be on the DEBUG reporting level, which is contradicting from my intention to make restoration more visible to users by default. Hence I'm proposing it at the restore-thread level. With all those rationales in mind, I think there are two options on the high level that we can provide: 1) provide the finer-grained level metrics from Streams i.e. task-level (the reason I did not consider store-level is that the internal physical state stores may evolve and changed), and let users to do their aggregations themselves if they like; but this require users to first turn on DEBUG level reporting being aware it's going to increasing the metrics reporting overhead; 2) provide the coarse-grained level metrics from Streams on the thread-level, and just get users to watch on that coarse-grained level. After some thought about that, my current preference is to split the group of metrics, and report rate metrics on the thread level, while reporting remaining and total num.records on the task level, but keeping the remaining.num.records at INFO level (btw just to answer Bruno's question, reporting remaining num.records does not require additional requests since we are already sending committed / log-end requests periodically to consider when we should complete / pause the restoration). 2) At the restoration thread, it would only either try to restore active tasks, or update standby tasks, with a priority to do the first as long as there are still active tasks needing restoration, at any given time. And hence I was originally thinking that we only need one rate metric, since if we split it then only one of them would ever have a non-zero value anyways. But after some thought I think it's still okay to split the metric since we do not have a state indicating if we are in "active-restore" or "standby-update" state anyways, and having those two metrics where only one of them would be non-zero would be valuable for users to tell when the state switched inside the thread. As a result, the proposed metrics would look like this: Thread-level (all INFO level): active-restoring-tasks standby-updating-tasks active-paused-tasks standby-paused-tasks idle-ratio restore-ratio checkpoint-ratio restore-call-rate active-restore-records-rate standby-update-records-rate Task-level: restored-total (DEBUG, as to align with process-total DEBUG) restored-rate (DEBUG, as to align with process-rate DEBUG) remaining-restore-records (INFO) Please let me know what you think. Guozhang On Mon, Oct 10, 2022 at 3:53 AM Nick Telford <nick.telf...@gmail.com> wrote: > Hi Guozhang, > > The metrics you've described are currently "thread-level" metrics. I know > you're planning on splitting out the restoration from the StreamThread, so > I question the utility of some of these metrics being thread-level, as they > won't correlate with other thread-level metrics, which correspond to a > StreamThread. > > Perhaps the record counting metrics would be more useful as > task/store-level metrics? That way, they can be aggregated by users to > determine things like total records remaining to restore by store, across > the entire app etc. > > Regards, > > Nick > > On Thu, 22 Sept 2022 at 13:21, Bruno Cadonna <cado...@apache.org> wrote: > > > Hi Guozhang! > > > > Thanks for the updates! > > > > 1. > > Why do you distinguish between active and standby for the total amount > > of restored/updated records but not for the rate of restored/updated > > records? > > > > 2. > > Regarding "standby-records-remaining" I am not sure how useful this > > metric is and I am not sure how hard it will be to record. I see the > > usefulness of monitoring the lag of a single standby state store, but I > > am not sure how useful it is to monitor the "lag" of a state updater > > thread that might potentially contain multiple state stores. > > Furthermore, do we need to issue a remote call to record this metric or > > do we get this information from each poll? > > > > 3. > > Could you please give an example where "active-records-restored-total" > > and "standby-records-updated-total" are useful? > > > > Best, > > Bruno > > > > On 20.09.22 22:45, Guozhang Wang wrote: > > > Hello Bruno/Nick, > > > > > > I've updated the KIP wiki to reflect the incorporated comments from > you, > > > please feel free to take another look and let me know what you think. > > > > > > > > > Guozhang > > > > > > On Tue, Sep 20, 2022 at 9:37 AM Guozhang Wang <wangg...@gmail.com> > > wrote: > > > > > >> Hi Nick, > > >> > > >> Thanks for the reviews, and I think these are good suggestions. Note > > that > > >> currently the `restore-records-total/rate` would include both > restoring > > >> active tasks as well as updating standby tasks, I think for your > > purposes > > >> you'd be more interested in active restoring tasks since they could > > >> complete, while updating standby tasks would not complete even if they > > have > > >> caught up with the active. At the same time, the changelog reader > would > > >> only be restoring either active or standby at a given time, and active > > >> tasks has a higher priority such that as long as there is at least one > > >> active task still restoring, we would not restore any standby tasks. > > From > > >> your suggestion, I'm thinking that maybe I should break up the rate / > > total > > >> metric for active and standby tasks separately. > > >> > > >> For deriving estimated time remaining though, the `total` metric may > not > > >> be helpful since they will not be "reset" after rebalances, i.e. they > > will > > >> be an ever-increasing number and record the total number of records > for > > the > > >> lifetime of the app. But still, just the remaining records alone, with > > the > > >> time elapsed monitored by the apps, should be sufficient to get the > > >> estimated time remaining. > > >> > > >> > > >> Guozhang > > >> > > >> > > >> On Tue, Sep 20, 2022 at 3:10 AM Nick Telford <nick.telf...@gmail.com> > > >> wrote: > > >> > > >>> Hi Guozhang, > > >>> > > >>> KIP looks great, I have one suggestion: in addition to > > >>> "restore-records-total", it would also be useful to track the number > of > > >>> records *remaining*, that is, the records that have not yet been > > restored. > > >>> This is actually the metric I was attempting to implement in the > > >>> StateRestoreListener that bumped me up against KAFKA-10575 :-) > > >>> > > >>> With both a "restore-records-total" and a "restore-remaining-total" > (or > > >>> similar) metric, it's possible to derive useful information like the > > >>> estimated time remaining for restoration (by dividing the remaining > > total > > >>> by the restoration rate). > > >>> > > >>> Regards, > > >>> > > >>> Nick > > >>> > > >>> On Mon, 19 Sept 2022 at 19:57, Guozhang Wang <wangg...@gmail.com> > > wrote: > > >>> > > >>>> Hello Bruno, > > >>>> > > >>>> Thanks for your comments! > > >>>> > > >>>> 1. Regarding the metrics group name: originally I put > > >>>> "stream-state-metrics" as it's related to state store restorations, > > but > > >>>> after a second thought I think I agree with you that this is quite > > >>>> confusing and not right. About the metrics groups, right now I have > > two > > >>>> ideas: > > >>>> > > >>>> a) Still use the metric group name "stream-thread-metrics", but > > >>>> differentiate with the processing threads on the thread id. The pros > > is > > >>>> that we do not introduce a new group, the cons is that users who > want > > to > > >>>> separate processing from restoration/updating in the future needs to > > do > > >>>> that on the thread id labels. > > >>>> b) Introduce a new group name, for example > > >>> "stream-state-updater-metrics" > > >>>> and still have a thread-id label. We would be introducing a new > group > > >>> which > > >>>> still have a thread-id, which may sound a bit weird (maybe if we do > > >>> that we > > >>>> could change the existing "stream-thread-metrics" into > > >>>> "stream-processor-metrics"). > > >>>> > > >>>> Right now I'm leaning towards b) and maybe in the future rename > > >>>> "thread-metrics" to "processor-metrics", LMK what do you think. > > >>>> > > >>>> 2. Regarding the metric names: today we may also pause a standby > > tasks, > > >>> and > > >>>> hence I'm trying to differentiate updating standbys from paused > > >>> standbys. > > >>>> Right now I'm thinking we can change "restoring-standby-tasks" to > > >>>> "updating-standby-tasks", and change all other metrics' "restore" > > >>> (except > > >>>> the "restoring-active-tasks") to "state-update", a.k.a > > >>>> "state-update-ratio", "state-update-records-total", > > >>>> "updating-standby-tasks" etc. > > >>>> > > >>>> 3. Regarding the function name: yeah I think that's a valid > concern, I > > >>> can > > >>>> change it to "onRestoreSuspended" since "Aborted" may confuse people > > >>> that > > >>>> previously called "onBatchRestored" are undone as part of the > > abortion. > > >>>> > > >>>> > > >>>> Guozhang > > >>>> > > >>>> > > >>>> > > >>>> On Mon, Sep 19, 2022 at 10:47 AM Bruno Cadonna <cado...@apache.org> > > >>> wrote: > > >>>> > > >>>>> Hi Guozhang, > > >>>>> > > >>>>> Thanks for the KIP! I think this KIP is a really nice addition to > > >>> better > > >>>>> understand what is going on in a Kafka Streams application. > > >>>>> > > >>>>> 1. > > >>>>> The metric names "paused-active-tasks" and "paused-standby-tasks" > > >>> might > > >>>>> be a bit confusing since at least active tasks can be paused also > > >>>>> outside of restoration. > > >>>>> > > >>>>> 2. > > >>>>> Why is the type of the metrics "stream-state-metrics"? I would have > > >>>>> expected "stream-thread-metrics" as the type. > > >>>>> > > >>>>> 3. > > >>>>> Isn't the value of the metric "restoring-standby-tasks" simply the > > >>>>> number of standby tasks since standby tasks are basically always > > >>>>> updating (aka restoring)? > > >>>>> > > >>>>> 4. > > >>>>> "idle-ratio", "restore-ratio", and "checkpoint-ratio" seem metrics > > >>>>> tailored to the upcoming state updater. They do not make much sense > > >>> with > > >>>>> a stream thread. Would it be better to introduce a new metrics > level > > >>>>> specifically for the state updater? > > >>>>> > > >>>>> 5. > > >>>>> Personally, I do not like to use the word "restoration" together > with > > >>>>> standbys since restoration somehow implies that there is an offset > > for > > >>>>> which the active task is considered restored and active processing > > can > > >>>>> start. In other words, restoration is finite. Standby tasks rather > > >>>>> update continuously their states. They can be up-to-date or > lagging. > > I > > >>>>> see that you could say "restored" instead of "up-to-date" and "not > > >>>>> restored" instead of "lagging", but IMO it does not describe well > the > > >>>>> situation. That is a rather minor point. I just wanted to mention > it. > > >>>>> > > >>>>> 6. > > >>>>> The name "onRestorePaused()" might be confusing since in Kafka > > Streams > > >>>>> users can also pause tasks. What about "onRestoreAborted()" or > > >>>>> "onRestoreSuspended"? > > >>>>> > > >>>>> Best, > > >>>>> Bruno > > >>>>> > > >>>>> > > >>>>> On 16.09.22 19:33, Guozhang Wang wrote: > > >>>>>> Hello everyone, > > >>>>>> > > >>>>>> I'd like to start a discussion for the following KIP, aiming to > > >>> improve > > >>>>>> Kafka Stream's restoration visibility via new metrics and callback > > >>>>> methods: > > >>>>>> > > >>>>>> > > >>>>> > > >>>> > > >>> > > > https://cwiki.apache.org/confluence/display/KAFKA/KIP-869%3A+Improve+Streams+State+Restoration+Visibility > > >>>>>> > > >>>>>> > > >>>>>> Thanks! > > >>>>>> -- Guozhang > > >>>>>> > > >>>>> > > >>>> > > >>>> > > >>>> -- > > >>>> -- Guozhang > > >>>> > > >>> > > >> > > >> > > >> -- > > >> -- Guozhang > > >> > > > > > > > > > -- -- Guozhang
