kafka-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ray Chiang <rchi...@apache.org>
Subject Re: [DISCUSS] KIP-346 - Limit blast radius of log compaction failure
Date Thu, 26 Jul 2018 20:19:06 GMT
Thanks for creating this KIP Stanislav.  My observations:

1) I agree with Colin that threads automatically re-launching threads 
generally isn't a great idea.  Metrics and/or monitoring threads are 
generally much safer.  And there's always the issue of what happens if 
the re-launcher dies?

2) I'm 100% with James in agreement with setting up the LogCleaner to 
skip over problematic partitions instead of dying.

3) There's a lot of "feature bloat" suggestions.  From how I see things, 
a message could get corrupted in one of several states:

3a) Message is corrupted by the leader partition saving to disk. 
Replicas have the same error.
3b) Message is corrupted by one of the replica partitions saving to 
disk.  Leader and other replica(s) unlikely to have the same error
3c) Disk corruption happens later (e.g. during partition move)

If we have the simplest solution, then all of the above will not cause 
the LogCleaner to crash and 3b/3c have a chance of manual recovery.

4) In most of the issues I'm seeing via work, most of the corruption 
seems persistent on the same log segment (i.e. a 3b/3c type of 
corruption).  The only improvement I can think of is that if such an 
error occurs, then have the option (configuration setting?) to create a 
<log_segment>.skip file (or something similar).  If the .skip file is 
there, don't re-scan the segment.  If you want a re-try or manage to fix 
the issue manually (e.g. copying from a replica), then the .skip file 
can be deleted after the segment is fixed and the LogCleaner will try 
again on the next iteration.

5) I'm in alignment with Colin's comment about hard drive failures. By 
the time you can reliably detect HDD hardware failures, it's less about 
improving the LogCleaner as much as that data needs to be moved to a new 


On 7/25/18 11:55 AM, Dhruvil Shah wrote:
> For the cleaner thread specifically, I do not think respawning will help at
> all because we are more than likely to run into the same issue again which
> would end up crashing the cleaner. Retrying makes sense for transient
> errors or when you believe some part of the system could have healed
> itself, both of which I think are not true for the log cleaner.
> On Wed, Jul 25, 2018 at 11:08 AM Ron Dagostino <rndgstn@gmail.com> wrote:
>> <<<respawning threads is likely to make things worse, by putting you in
>> infinite loop which consumes resources and fires off continuous log
>> messages.
>> Hi Colin.  In case it could be relevant, one way to mitigate this effect is
>> to implement a backoff mechanism (if a second respawn is to occur then wait
>> for 1 minute before doing it; then if a third respawn is to occur wait for
>> 2 minutes before doing it; then 4 minutes, 8 minutes, etc. up to some max
>> wait time).
>> I have no opinion on whether respawn is appropriate or not in this context,
>> but a mitigation like the increasing backoff described above may be
>> relevant in weighing the pros and cons.
>> Ron
>> On Wed, Jul 25, 2018 at 1:26 PM Colin McCabe <cmccabe@apache.org> wrote:
>>> On Mon, Jul 23, 2018, at 23:20, James Cheng wrote:
>>>> Hi Stanislav! Thanks for this KIP!
>>>> I agree that it would be good if the LogCleaner were more tolerant of
>>>> errors. Currently, as you said, once it dies, it stays dead.
>>>> Things are better now than they used to be. We have the metric
>>>>        kafka.log:type=LogCleanerManager,name=time-since-last-run-ms
>>>> which we can use to tell us if the threads are dead. And as of 1.1.0,
>> we
>>>> have KIP-226, which allows you to restart the log cleaner thread,
>>>> without requiring a broker restart.
>> https://cwiki.apache.org/confluence/display/KAFKA/KIP-226+-+Dynamic+Broker+Configuration
>>>> <
>> https://cwiki.apache.org/confluence/display/KAFKA/KIP-226+-+Dynamic+Broker+Configuration
>>>> I've only read about this, I haven't personally tried it.
>>> Thanks for pointing this out, James!  Stanislav, we should probably add a
>>> sentence or two mentioning the KIP-226 changes somewhere in the KIP.
>> Maybe
>>> in the intro section?
>>> I think it's clear that requiring the users to manually restart the log
>>> cleaner is not a very good solution.  But it's good to know that it's a
>>> possibility on some older releases.
>>>> Some comments:
>>>> * I like the idea of having the log cleaner continue to clean as many
>>>> partitions as it can, skipping over the problematic ones if possible.
>>>> * If the log cleaner thread dies, I think it should automatically be
>>>> revived. Your KIP attempts to do that by catching exceptions during
>>>> execution, but I think we should go all the way and make sure that a
>> new
>>>> one gets created, if the thread ever dies.
>>> This is inconsistent with the way the rest of Kafka works.  We don't
>>> automatically re-create other threads in the broker if they terminate.
>> In
>>> general, if there is a serious bug in the code, respawning threads is
>>> likely to make things worse, by putting you in an infinite loop which
>>> consumes resources and fires off continuous log messages.
>>>> * It might be worth trying to re-clean the uncleanable partitions. I've
>>>> seen cases where an uncleanable partition later became cleanable. I
>>>> unfortunately don't remember how that happened, but I remember being
>>>> surprised when I discovered it. It might have been something like a
>>>> follower was uncleanable but after a leader election happened, the log
>>>> truncated and it was then cleanable again. I'm not sure.
>>> James, I disagree.  We had this behavior in the Hadoop Distributed File
>>> System (HDFS) and it was a constant source of user problems.
>>> What would happen is disks would just go bad over time.  The DataNode
>>> would notice this and take them offline.  But then, due to some
>>> "optimistic" code, the DataNode would periodically try to re-add them to
>>> the system.  Then one of two things would happen: the disk would just
>> fail
>>> immediately again, or it would appear to work and then fail after a short
>>> amount of time.
>>> The way the disk failed was normally having an I/O request take a really
>>> long time and time out.  So a bunch of request handler threads would
>>> basically slam into a brick wall when they tried to access the bad disk,
>>> slowing the DataNode to a crawl.  It was even worse in the second
>> scenario,
>>> if the disk appeared to work for a while, but then failed.  Any data that
>>> had been written on that DataNode to that disk would be lost, and we
>> would
>>> need to re-replicate it.
>>> Disks aren't biological systems-- they don't heal over time.  Once
>> they're
>>> bad, they stay bad.  The log cleaner needs to be robust against cases
>> where
>>> the disk really is failing, and really is returning bad data or timing
>> out.
>>>> * For your metrics, can you spell out the full metric in JMX-style
>>>> format, such as:
>>   kafka.log:type=LogCleanerManager,name=uncleanable-partitions-count
>>>>                value=4
>>>> * For "uncleanable-partitions": topic-partition names can be very long.
>>>> I think the current max size is 210 characters (or maybe 240-ish?).
>>>> Having the "uncleanable-partitions" being a list could be very large
>>>> metric. Also, having the metric come out as a csv might be difficult to
>>>> work with for monitoring systems. If we *did* want the topic names to
>> be
>>>> accessible, what do you think of having the
>>>>        kafka.log:type=LogCleanerManager,topic=topic1,partition=2
>>>> I'm not sure if LogCleanerManager is the right type, but my example was
>>>> that the topic and partition can be tags in the metric. That will allow
>>>> monitoring systems to more easily slice and dice the metric. I'm not
>>>> sure what the attribute for that metric would be. Maybe something like
>>>> "uncleaned bytes" for that topic-partition? Or time-since-last-clean?
>> Or
>>>> maybe even just "Value=1".
>>> I haven't though about this that hard, but do we really need the
>>> uncleanable topic names to be accessible through a metric?  It seems like
>>> the admin should notice that uncleanable partitions are present, and then
>>> check the logs?
>>>> * About `max.uncleanable.partitions`, you said that this likely
>>>> indicates that the disk is having problems. I'm not sure that is the
>>>> case. For the 4 JIRAs that you mentioned about log cleaner problems,
>> all
>>>> of them are partition-level scenarios that happened during normal
>>>> operation. None of them were indicative of disk problems.
>>> I don't think this is a meaningful comparison.  In general, we don't
>>> accept JIRAs for hard disk problems that happen on a particular cluster.
>>> If someone opened a JIRA that said "my hard disk is having problems" we
>>> could close that as "not a Kafka bug."  This doesn't prove that disk
>>> problems don't happen, but  just that JIRA isn't the right place for
>> them.
>>> I do agree that the log cleaner has had a significant number of logic
>>> bugs, and that we need to be careful to limit their impact.  That's one
>>> reason why I think that a threshold of "number of uncleanable logs" is a
>>> good idea, rather than just failing after one IOException.  In all the
>>> cases I've seen where a user hit a logic bug in the log cleaner, it was
>>> just one partition that had the issue.  We also should increase test
>>> coverage for the log cleaner.
>>>> * About marking disks as offline when exceeding a certain threshold,
>>>> that actually increases the blast radius of log compaction failures.
>>>> Currently, the uncleaned partitions are still readable and writable.
>>>> Taking the disks offline would impact availability of the uncleanable
>>>> partitions, as well as impact all other partitions that are on the
>> disk.
>>> In general, when we encounter I/O errors, we take the disk partition
>>> offline.  This is spelled out in KIP-112 (
>> https://cwiki.apache.org/confluence/display/KAFKA/KIP-112%3A+Handle+disk+failure+for+JBOD
>>> ) :
>>>> - Broker assumes a log directory to be good after it starts, and mark
>>> log directory as
>>>> bad once there is IOException when broker attempts to access (i.e. read
>>> or write) the log directory.
>>>> - Broker will be offline if all log directories are bad.
>>>> - Broker will stop serving replicas in any bad log directory. New
>>> replicas will only be created
>>>> on good log directory.
>>> The behavior Stanislav is proposing for the log cleaner is actually more
>>> optimistic than what we do for regular broker I/O, since we will tolerate
>>> multiple IOExceptions, not just one.  But it's generally consistent.
>>> Ignoring errors is not.  In any case, if you want to tolerate an
>> unlimited
>>> number of I/O errors, you can just set the threshold to an infinite value
>>> (although I think that would be a bad idea).
>>> best,
>>> Colin
>>>> -James
>>>>> On Jul 23, 2018, at 5:46 PM, Stanislav Kozlovski <
>>> stanislav@confluent.io> wrote:
>>>>> I renamed the KIP and that changed the link. Sorry about that. Here
>> is
>>> the
>>>>> new link:
>> https://cwiki.apache.org/confluence/display/KAFKA/KIP-346+-+Improve+LogCleaner+behavior+on+error
>>>>> On Mon, Jul 23, 2018 at 5:11 PM Stanislav Kozlovski <
>>> stanislav@confluent.io>
>>>>> wrote:
>>>>>> Hey group,
>>>>>> I created a new KIP about making log compaction more fault-tolerant.
>>>>>> Please give it a look here and please share what you think,
>>> especially in
>>>>>> regards to the points in the "Needs Discussion" paragraph.
>>>>>> KIP: KIP-346
>>>>>> <
>> https://cwiki.apache.org/confluence/display/KAFKA/KIP-346+-+Limit+blast+radius+of+log+compaction+failure
>>>>>> --
>>>>>> Best,
>>>>>> Stanislav
>>>>> --
>>>>> Best,
>>>>> Stanislav

View raw message