[jira] [Commented] (MAPREDUCE-7282) MR v2 commit algorithm should be deprecated and not the default
[ https://issues.apache.org/jira/browse/MAPREDUCE-7282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17209768#comment-17209768 ] Steve Loughran commented on MAPREDUCE-7282: --- Created MAPREDUCE-7300 which adds ability to ask a committer whether failures during task attempt commit are recoverable. combined with configurable policy on what to do here: warn + continue vs fail, people can stick with v2 committer but can be happy that if something went wrong during task commit, they'd know. > MR v2 commit algorithm should be deprecated and not the default > --- > > Key: MAPREDUCE-7282 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-7282 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2 >Affects Versions: 3.3.0, 3.2.1, 3.1.3, 3.3.1 >Reporter: Steve Loughran >Priority: Major > > The v2 MR commit algorithm moves files from the task attempt dir into the > dest dir on task commit -one by one > It is therefore not atomic > # if a task commit fails partway through and another task attempt commits > -unless exactly the same filenames are used, output of the first attempt may > be included in the final result > # if a worker partitions partway through task commit, and then continues > after another attempt has committed, it may partially overwrite the output > -even when the filenames are the same > Both MR and spark assume that task commits are atomic. Either they need to > consider that this is not the case, we add a way to probe for a committer > supporting atomic task commit, and the engines both add handling for task > commit failures (probably fail job) > Better: we remove this as the default, maybe also warn when it is being used -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: mapreduce-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: mapreduce-issues-h...@hadoop.apache.org
[jira] [Commented] (MAPREDUCE-7282) MR v2 commit algorithm should be deprecated and not the default
[ https://issues.apache.org/jira/browse/MAPREDUCE-7282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17208145#comment-17208145 ] Steve Loughran commented on MAPREDUCE-7282: --- bq. Tasks request permission from the AM to commit. yes, and then we assume that they continue to completion, rather than pausing for an extended period of time, so by the time the AM/spark driver gets a timeout, it can be assumed to be one of a network failure or the worker has failed/VM/k8s container terminated. The "suspended for a long time and then continues" risk does exist, and is unlikely on a physical cluster, but in a world of VMs, not entirely inconceivable. I note the MR AM does track its time from last heartbeat to the YARN RM to detect partitions, workers don't. > MR v2 commit algorithm should be deprecated and not the default > --- > > Key: MAPREDUCE-7282 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-7282 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2 >Affects Versions: 3.3.0, 3.2.1, 3.1.3, 3.3.1 >Reporter: Steve Loughran >Priority: Major > > The v2 MR commit algorithm moves files from the task attempt dir into the > dest dir on task commit -one by one > It is therefore not atomic > # if a task commit fails partway through and another task attempt commits > -unless exactly the same filenames are used, output of the first attempt may > be included in the final result > # if a worker partitions partway through task commit, and then continues > after another attempt has committed, it may partially overwrite the output > -even when the filenames are the same > Both MR and spark assume that task commits are atomic. Either they need to > consider that this is not the case, we add a way to probe for a committer > supporting atomic task commit, and the engines both add handling for task > commit failures (probably fail job) > Better: we remove this as the default, maybe also warn when it is being used -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: mapreduce-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: mapreduce-issues-h...@hadoop.apache.org
[jira] [Commented] (MAPREDUCE-7282) MR v2 commit algorithm should be deprecated and not the default
[
https://issues.apache.org/jira/browse/MAPREDUCE-7282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17204264#comment-17204264
]
Daryn Sharp commented on MAPREDUCE-7282:
I'm also -1 on changing the default. It exposes users to new (old but new to
them) behavior that may have quirks. This was a 2.7 change from 5 years ago so
if it's a high risk issue our customers would have squawked by now. Has this
been frequently observed or theorized?
Notably our users won't tolerate the performance regression and SLA misses. I
seem to recall jobs that ran for a single-digit minutes followed by a
double-digit commit. The v2 commit amortized the commit to under a minute.
I'm not a MR expert. Here's my understanding:
{quote}if a task commit fails partway through and another task attempt commits
-unless exactly the same filenames are used, output of the first attempt may be
included in the final result
{quote}
Isn't that indicative of a non-deterministic job? Should the risk to a few
"bad" jobs outweigh the benefit to the mass majority of jobs? Why not change
the committer for at risk jobs?
{quote}if a worker partitions partway through task commit, and then continues
after another attempt has committed, it may partially overwrite the output
-even when the filenames are the same
{quote}
I don't think this can happen. Tasks request permission from the AM to commit.
> MR v2 commit algorithm should be deprecated and not the default
> ---
>
> Key: MAPREDUCE-7282
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-7282
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2
>Affects Versions: 3.3.0, 3.2.1, 3.1.3, 3.3.1
>Reporter: Steve Loughran
>Priority: Major
>
> The v2 MR commit algorithm moves files from the task attempt dir into the
> dest dir on task commit -one by one
> It is therefore not atomic
> # if a task commit fails partway through and another task attempt commits
> -unless exactly the same filenames are used, output of the first attempt may
> be included in the final result
> # if a worker partitions partway through task commit, and then continues
> after another attempt has committed, it may partially overwrite the output
> -even when the filenames are the same
> Both MR and spark assume that task commits are atomic. Either they need to
> consider that this is not the case, we add a way to probe for a committer
> supporting atomic task commit, and the engines both add handling for task
> commit failures (probably fail job)
> Better: we remove this as the default, maybe also warn when it is being used
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: mapreduce-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: mapreduce-issues-h...@hadoop.apache.org
[jira] [Commented] (MAPREDUCE-7282) MR v2 commit algorithm should be deprecated and not the default
[ https://issues.apache.org/jira/browse/MAPREDUCE-7282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17204156#comment-17204156 ] Steve Loughran commented on MAPREDUCE-7282: --- [~Jim_Brennan] -have a look at the latest PR. this retains it, simply changes the default and logs @ WARN When you use V1...and uses a special log for that warning so you can turn it off without running the risk of losing important messages > MR v2 commit algorithm should be deprecated and not the default > --- > > Key: MAPREDUCE-7282 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-7282 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2 >Affects Versions: 3.3.0, 3.2.1, 3.1.3, 3.3.1 >Reporter: Steve Loughran >Priority: Major > > The v2 MR commit algorithm moves files from the task attempt dir into the > dest dir on task commit -one by one > It is therefore not atomic > # if a task commit fails partway through and another task attempt commits > -unless exactly the same filenames are used, output of the first attempt may > be included in the final result > # if a worker partitions partway through task commit, and then continues > after another attempt has committed, it may partially overwrite the output > -even when the filenames are the same > Both MR and spark assume that task commits are atomic. Either they need to > consider that this is not the case, we add a way to probe for a committer > supporting atomic task commit, and the engines both add handling for task > commit failures (probably fail job) > Better: we remove this as the default, maybe also warn when it is being used -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: mapreduce-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: mapreduce-issues-h...@hadoop.apache.org
[jira] [Commented] (MAPREDUCE-7282) MR v2 commit algorithm should be deprecated and not the default
[
https://issues.apache.org/jira/browse/MAPREDUCE-7282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17201605#comment-17201605
]
Hadoop QA commented on MAPREDUCE-7282:
--
| (x) *{color:red}-1 overall{color}* |
\\
\\
|| Vote || Subsystem || Runtime || Logfile || Comment ||
| {color:blue}0{color} | {color:blue} reexec {color} | {color:blue} 1m
15s{color} | {color:blue}{color} | {color:blue} Docker mode activated. {color} |
|| || || || {color:brown} Prechecks {color} || ||
| {color:green}+1{color} | {color:green} dupname {color} | {color:green} 0m
0s{color} | {color:green}{color} | {color:green} No case conflicting files
found. {color} |
| {color:green}+1{color} | {color:green} @author {color} | {color:green} 0m
0s{color} | {color:green}{color} | {color:green} The patch does not contain any
@author tags. {color} |
| {color:green}+1{color} | {color:green} {color} | {color:green} 0m 0s{color}
| {color:green}test4tests{color} | {color:green} The patch appears to include 2
new or modified test files. {color} |
|| || || || {color:brown} trunk Compile Tests {color} || ||
| {color:green}+1{color} | {color:green} mvninstall {color} | {color:green} 31m
21s{color} | {color:green}{color} | {color:green} trunk passed {color} |
| {color:green}+1{color} | {color:green} compile {color} | {color:green} 0m
39s{color} | {color:green}{color} | {color:green} trunk passed with JDK
Ubuntu-11.0.8+10-post-Ubuntu-0ubuntu118.04.1 {color} |
| {color:green}+1{color} | {color:green} compile {color} | {color:green} 0m
34s{color} | {color:green}{color} | {color:green} trunk passed with JDK Private
Build-1.8.0_265-8u265-b01-0ubuntu2~18.04-b01 {color} |
| {color:green}+1{color} | {color:green} checkstyle {color} | {color:green} 0m
29s{color} | {color:green}{color} | {color:green} trunk passed {color} |
| {color:green}+1{color} | {color:green} mvnsite {color} | {color:green} 0m
45s{color} | {color:green}{color} | {color:green} trunk passed {color} |
| {color:green}+1{color} | {color:green} shadedclient {color} | {color:green}
15m 41s{color} | {color:green}{color} | {color:green} branch has no errors when
building and testing our client artifacts. {color} |
| {color:green}+1{color} | {color:green} javadoc {color} | {color:green} 0m
25s{color} | {color:green}{color} | {color:green} trunk passed with JDK
Ubuntu-11.0.8+10-post-Ubuntu-0ubuntu118.04.1 {color} |
| {color:green}+1{color} | {color:green} javadoc {color} | {color:green} 0m
23s{color} | {color:green}{color} | {color:green} trunk passed with JDK Private
Build-1.8.0_265-8u265-b01-0ubuntu2~18.04-b01 {color} |
| {color:blue}0{color} | {color:blue} spotbugs {color} | {color:blue} 1m
15s{color} | {color:blue}{color} | {color:blue} Used deprecated FindBugs
config; considering switching to SpotBugs. {color} |
| {color:green}+1{color} | {color:green} findbugs {color} | {color:green} 1m
13s{color} | {color:green}{color} | {color:green} trunk passed {color} |
|| || || || {color:brown} Patch Compile Tests {color} || ||
| {color:green}+1{color} | {color:green} mvninstall {color} | {color:green} 0m
30s{color} | {color:green}{color} | {color:green} the patch passed {color} |
| {color:green}+1{color} | {color:green} compile {color} | {color:green} 0m
32s{color} | {color:green}{color} | {color:green} the patch passed with JDK
Ubuntu-11.0.8+10-post-Ubuntu-0ubuntu118.04.1 {color} |
| {color:green}+1{color} | {color:green} javac {color} | {color:green} 0m
32s{color} | {color:green}{color} | {color:green} the patch passed {color} |
| {color:green}+1{color} | {color:green} compile {color} | {color:green} 0m
28s{color} | {color:green}{color} | {color:green} the patch passed with JDK
Private Build-1.8.0_265-8u265-b01-0ubuntu2~18.04-b01 {color} |
| {color:green}+1{color} | {color:green} javac {color} | {color:green} 0m
28s{color} | {color:green}{color} | {color:green} the patch passed {color} |
| {color:orange}-0{color} | {color:orange} checkstyle {color} | {color:orange}
0m 20s{color} |
{color:orange}https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-2320/3/artifact/out/diff-checkstyle-hadoop-mapreduce-project_hadoop-mapreduce-client_hadoop-mapreduce-client-core.txt{color}
| {color:orange}
hadoop-mapreduce-project/hadoop-mapreduce-client/hadoop-mapreduce-client-core:
The patch generated 1 new + 46 unchanged - 5 fixed = 47 total (was 51) {color} |
| {color:green}+1{color} | {color:green} mvnsite {color} | {color:green} 0m
30s{color} | {color:green}{color} | {color:green} the patch passed {color} |
| {color:green}+1{color} | {color:green} whitespace {color} | {color:green} 0m
0s{color} | {color:green}{color} | {color:green} The patch has no whitespace
issues. {color} |
| {color:green}+1{color} | {color:green} xml {color} | {color:green} 0m
2s{color} | {color:green}{color} | {color:green} The patch has no ill-formed
XML file. {color} |
| {color:green}+1{color} |
[jira] [Commented] (MAPREDUCE-7282) MR v2 commit algorithm should be deprecated and not the default
[ https://issues.apache.org/jira/browse/MAPREDUCE-7282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17201046#comment-17201046 ] Jim Brennan commented on MAPREDUCE-7282: I am -1 on the proposal to remove the v2 algorithm. Please see [~jlowe]'s [comment|https://issues.apache.org/jira/browse/MAPREDUCE-4815?focusedCommentId=14271115=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14271115] from the original discussion on [MAPREDUCE-4815]. We have been running with the v2 algorithm in production on large clusters for years at Verizon Media (Yahoo). I don't think it is appropriate to remove it. > MR v2 commit algorithm should be deprecated and not the default > --- > > Key: MAPREDUCE-7282 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-7282 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2 >Affects Versions: 3.3.0, 3.2.1, 3.1.3, 3.3.1 >Reporter: Steve Loughran >Priority: Major > > The v2 MR commit algorithm moves files from the task attempt dir into the > dest dir on task commit -one by one > It is therefore not atomic > # if a task commit fails partway through and another task attempt commits > -unless exactly the same filenames are used, output of the first attempt may > be included in the final result > # if a worker partitions partway through task commit, and then continues > after another attempt has committed, it may partially overwrite the output > -even when the filenames are the same > Both MR and spark assume that task commits are atomic. Either they need to > consider that this is not the case, we add a way to probe for a committer > supporting atomic task commit, and the engines both add handling for task > commit failures (probably fail job) > Better: we remove this as the default, maybe also warn when it is being used -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: mapreduce-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: mapreduce-issues-h...@hadoop.apache.org
