Awesome, what about with the fix, but with the default list of hashing
algos (SHA-512, SHA-256, SHA-1, MD5) in place?
Am 2020-09-27 um 00:07 schrieb Dan Tran:
it works using the provided commitID +
-Daether.checksums.algorithms=SHA1,MD5
-D
On Sat, Sep 26, 2020 at 2:25 PM Michael Osipov <micha...@apache.org>
wrote:
Gosh, I assumed so. Every SHA family is horrible in terms of
performance. This is going to be a huge problem for the community, I
would prefer Blake2 or Blake3 for this.
Anyway, I think this is the root cause, actually the fix:
https://github.com/apache/maven-resolver/commit/a9e0ca8307e532fc30591a47d7d96e27fa647465
Can you please try with the property set
(-Daether.checksums.algorithms=SHA1,MD5)?
Thrilled to see your results. If this works, please remove the property
and see whether the patch along with SHA-2 on URLs changes the
performance as well.
Michael
Am 2020-09-26 um 21:45 schrieb Dan Tran:
I am able to pin down the problematic commit:
https://github.com/apache/maven-resolver/commit/8e8a4f1944a0589397e36a5dd049ec765424e863
Adding -Daether.checksums.algorithms=SHA1,MD5 to maven command line
does
not help
Sorry about the delay
-D
On Wed, Sep 16, 2020 at 1:57 AM Michael Osipov <micha...@apache.org>
wrote:
Your logging config is sound, so appearently nothing is wrong.
You see no trace because DefaultSyncContextFactory is empty and there
is
nothing to trace. Something else is the cause.
Can you git bisect 1.4.2 and 1.6.0 to find the offending commit?
Otherwise I have no idea and we are searching the needle in the
haystack.
The only commits which come to my mind are:
*
https://github.com/apache/maven-resolver/commit/8e8a4f1944a0589397e36a5dd049ec765424e863?w=1
You can disable/restore this one by using:
-Daether.checksums.algorithms=SHA1,MD5 (previous behavior)
*
https://github.com/apache/maven-resolver/commit/3c2a5141c9d393a45bdee873e6aa2c9c275b7b58?w=1
But still this wouldn't make any sense.
Looking forward for some findings.
Am 2020-09-15 um 22:45 schrieb Dan Tran:
this is my log config
org.slf4j.simpleLogger.defaultLogLevel=info
org.slf4j.simpleLogger.showDateTime=true
org.slf4j.simpleLogger.showThreadName=true
org.slf4j.simpleLogger.showShortLogName=true
org.slf4j.simpleLogger.showLogName=false
org.slf4j.simpleLogger.logFile=System.out
org.slf4j.simpleLogger.cacheOutputStream=true
org.slf4j.simpleLogger.levelInBrackets=true
org.slf4j.simpleLogger.log.Sisu=info
org.slf4j.simpleLogger.warnLevelString=WARNING
# MNG-6181: mvn -X also prints all debug logging from HttpClient
# Be aware that the shaded packages are used
# org.apache.http -> org.apache.maven.wagon.providers.http.httpclient
org.slf4j.simpleLogger.log.org.apache.maven.wagon.providers.http.httpclient=off
org.slf4j.simpleLogger.log.org.apache.maven.wagon.providers.http.httpclient.wire=off
org.slf4j.simpleLogger.log.org.eclipse.aether.synccontext=trace
no sign of [TRACE]
On Tue, Sep 15, 2020 at 1:30 PM Dan Tran <dant...@gmail.com> wrote:
Maven 3.7.0 with Resolver 1.4.2 - no issue
Maven 3.7.0 with Resolver 1.6.0 + trace. I cant get any indication
about
4626 [main] [TRACE] GlobalSyncContextFactory$GlobalSyncContext -
Acquiring global...
35318 [main] [TRACE] GlobalSyncContextFactory$GlobalSyncContext -
Releasing global...
Very strange
-D
On Tue, Sep 15, 2020 at 4:37 AM Michael Osipov <micha...@apache.org
wrote:
There is no new feature active in 1.6.0 in this regard.
Here is the diff:
https://github.com/apache/maven-resolver/compare/maven-resolver-1.4.2...maven-resolver-1.6.0?w=1
Since I do not have access to your project, can you please try the
following:
* Maven 3.7.0 with Resolver 1.4.2
* Maven 3.6.3 with Resolver 1.6.0
* If 3.7.0/1.4.2 does not affect the build, do a git bisect with
1.4.2
and 1.6.0.
Alternatively, provide the log output as configured here:
https://maven.apache.org/resolver-archives/resolver-LATEST/maven-resolver-synccontext-global/index.html#Debugging
In 1.6.0 there is even less synchronization because the
TrackingFileManager does not lock anything anymore.
I will hold off vote completion until I get some analyzable data
from
you.
Michael
Am 2020-09-15 um 11:49 schrieb Dan Tran:
I build maven 3.7.0 + MResolve 1.6.0. The same slow build
observed
at
my
300+ modules ( 60min versus 5min)
-1 ( non-binding) from me. there must and option to disable this
new
feature
-D
On Mon, Sep 14, 2020 at 4:33 AM Michael Osipov <
micha...@apache.org
wrote:
Am 2020-09-14 um 10:57 schrieb Dan Tran:
sorry about the fast finger, my test shows a huge increase in
build
time
for large builds. Possible to disable this feature?
This cannot be. It has been superceded by
https://issues.apache.org/jira/browse/MRESOLVER-130. The
GlobalSyncContextFactory have been moved to a separate module,
the
original behavior has been restored. I.e., this feature is not
enabled
by default.
Are you certain that you don't have leftovers in
${maven.home}/lib/ext?
On Mon, Sep 14, 2020 at 1:49 AM Dan Tran <dant...@gmail.com>
wrote:
possible to disable this feature
https://issues.apache.org/jira/browse/MRESOLVER-123 ?
On Mon, Sep 14, 2020 at 12:06 AM Romain Manni-Bucau <
rmannibu...@gmail.com>
wrote:
+1, tested in java 8 with mvn 3.6.3.
Romain Manni-Bucau
@rmannibucau <https://twitter.com/rmannibucau> | Blog
<https://rmannibucau.metawerx.net/> | Old Blog
<http://rmannibucau.wordpress.com> | Github <
https://github.com/rmannibucau> |
LinkedIn <https://www.linkedin.com/in/rmannibucau> | Book
<
https://www.packtpub.com/application-development/java-ee-8-high-performance
Le lun. 14 sept. 2020 à 08:58, Enrico Olivelli <
eolive...@gmail.com>
a
écrit :
+1
checked checksum and run tests on JDK14 + Mac
mvn -v
Apache Maven 3.6.3 (cecedd343002696d0abb50b32b541b8a6ba2883f)
Maven home: xxxxxx
Java version: 14, vendor: AdoptOpenJDK, runtime:
/Library/Java/JavaVirtualMachines/adoptopenjdk-14.jdk/Contents/Home
Default locale: it_IT, platform encoding: UTF-8
OS name: "mac os x", version: "10.15.4", arch: "x86_64",
family:
"mac"
Il giorno lun 14 set 2020 alle ore 08:55 Hervé BOUTEMY <
herve.bout...@free.fr> ha scritto:
+1
sadly, the build is not reproducible: reference build done
with
JDK 8
on
Windows, but Felix Maven Bundle Plugin is causing issues in
generated
META-
INF/MANIFEST.MF (contains username, detailed JDK version,
and
order
or
Private
Package not reproducible): I'll have a look to check if
recent
updates to
this
plugin can fix the issue...
Regards,
Hervé
Le vendredi 11 septembre 2020, 20:19:56 CEST Michael Osipov
a
écrit :
Hi,
We solved 19 issues:
https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12320628&ve
rsion=12348666
There are still a couple of issues left in JIRA:
https://issues.apache.org/jira/projects/MRESOLVER/issues?filter=allopenissue
s
Staging repo:
https://repository.apache.org/content/repositories/maven-1605/
https://repository.apache.org/content/repositories/maven-1605/org/apache/mav
en/resolver/maven-resolver/1.6.0/maven-resolver-1.6.0-source-release.zip
Source release checksum(s):
maven-resolver-1.6.0-source-release.zip
fd419a3f0bbe3c0ea0b3d7b881c4892fcfc73ec6cf812e7ddb724febb8bc010c754d5845498c
6c916a3bbdae2d90dd66787eb44fdbdd27310c745bdc1e8cc15b
Staging site:
https://maven.apache.org/resolver-archives/resolver-LATEST/
Guide to testing staged releases:
https://maven.apache.org/guides/development/guide-testing-releases.html
Note: This Resolver version offers salvation from a 13.5
years
old
feature request: MNG-2802
Yes, you can have now concurrent-safe access to your local
Maven
repository synchronized with Redis.
Vote open for 72 hours.
[ ] +1
[ ] +0
[ ] -1
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
For additional commands, e-mail: dev-h...@maven.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
For additional commands, e-mail: dev-h...@maven.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
For additional commands, e-mail: dev-h...@maven.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
For additional commands, e-mail: dev-h...@maven.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
For additional commands, e-mail: dev-h...@maven.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
For additional commands, e-mail: dev-h...@maven.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org
For additional commands, e-mail: dev-h...@maven.apache.org
