[jira] [Work logged] (ARTEMIS-4401) slow performance of JDBC while paging
[ https://issues.apache.org/jira/browse/ARTEMIS-4401?focusedWorklogId=878469=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-878469 ] ASF GitHub Bot logged work on ARTEMIS-4401: --- Author: ASF GitHub Bot Created on: 26/Aug/23 01:08 Start Date: 26/Aug/23 01:08 Worklog Time Spent: 10m Work Description: clebertsuconic commented on PR #4591: URL: https://github.com/apache/activemq-artemis/pull/4591#issuecomment-1694095633 I'm seeing this as draft for: - Renaming some tests - If you set page-size > 1MB performance will be pretty bad with JDBC - Some other minor things. Issue Time Tracking --- Worklog Id: (was: 878469) Time Spent: 20m (was: 10m) > slow performance of JDBC while paging > - > > Key: ARTEMIS-4401 > URL: https://issues.apache.org/jira/browse/ARTEMIS-4401 > Project: ActiveMQ Artemis > Issue Type: Bug >Reporter: Clebert Suconic >Assignee: Clebert Suconic >Priority: Major > Fix For: 2.31.0 > > Time Spent: 20m > Remaining Estimate: 0h > > Syncs are not properly implemented into page files. Basically ever message > sent and every blob update is issuing a sync update on the context and a > commit on the Database. > Certain databases will have an OK performance with lots of commits (e.g. > Postgres) but I'm not sure how correct is the blob update. > As part of this task I'm creating a few Database tests to validate these > changes. -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Work logged] (ARTEMIS-4349) Replace Guava cache with Caffeine
[ https://issues.apache.org/jira/browse/ARTEMIS-4349?focusedWorklogId=878428=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-878428 ] ASF GitHub Bot logged work on ARTEMIS-4349: --- Author: ASF GitHub Bot Created on: 25/Aug/23 15:17 Start Date: 25/Aug/23 15:17 Worklog Time Spent: 10m Work Description: gemmellr commented on PR #4584: URL: https://github.com/apache/activemq-artemis/pull/4584#issuecomment-1693527991 > > So I guess its just a case of deciding whether we want to retain the guava-style inline execution behaviour, or change to the Caffeine async execution default. > > Regarding the cache usage in `org.apache.activemq.artemis.core.security.impl.SecurityStoreImpl` my thoughts are: > > 1. The only use-case where it really matters is with size = 0 and that's been fixed that already via [ARTEMIS-4399 fix disabled authn/z cache #4589](https://github.com/apache/activemq-artemis/pull/4589). > > 2. It makes sense to optimize security since it is used basically every time a client connects. > > > Therefore, I'm in favor of using the default from Caffeine. Fair enough, since the SecurityStoreImpl doesnt look to use any of the listener stuff so it probably isn't really much/any different. The other usage does use it though, so I want to look into whether its prepared to handle the new concurrency that could result from offloading to the FJ pool (though if not, it may also not be prepared to handle any existing concurrency that could occur already.) Issue Time Tracking --- Worklog Id: (was: 878428) Time Spent: 10h (was: 9h 50m) > Replace Guava cache with Caffeine > - > > Key: ARTEMIS-4349 > URL: https://issues.apache.org/jira/browse/ARTEMIS-4349 > Project: ActiveMQ Artemis > Issue Type: Improvement >Affects Versions: 2.29.0 >Reporter: Alexey Markevich >Priority: Major > Time Spent: 10h > Remaining Estimate: 0h > > based on benchmark https://github.com/ben-manes/caffeine/wiki/Benchmarks -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Work logged] (ARTEMIS-4349) Replace Guava cache with Caffeine
[
https://issues.apache.org/jira/browse/ARTEMIS-4349?focusedWorklogId=878420=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-878420
]
ASF GitHub Bot logged work on ARTEMIS-4349:
---
Author: ASF GitHub Bot
Created on: 25/Aug/23 14:53
Start Date: 25/Aug/23 14:53
Worklog Time Spent: 10m
Work Description: gemmellr commented on code in PR #4584:
URL: https://github.com/apache/activemq-artemis/pull/4584#discussion_r1305770283
##
artemis-server/src/main/java/org/apache/activemq/artemis/core/security/impl/SecurityStoreImpl.java:
##
@@ -101,15 +103,15 @@ public SecurityStoreImpl(final
HierarchicalRepository> securityReposit
if (authenticationCacheSize == 0) {
authenticationCache = null;
} else {
- authenticationCache = CacheBuilder.newBuilder()
+ authenticationCache = Caffeine.newBuilder()
.maximumSize(authenticationCacheSize)
.expireAfterWrite(invalidationInterval, TimeUnit.MILLISECONDS)
.build();
Review Comment:
Fair. That was probably result of rebasing the changes again on top of your
earlier PR, and those lines not having changes on them.
Issue Time Tracking
---
Worklog Id: (was: 878420)
Time Spent: 9h 50m (was: 9h 40m)
> Replace Guava cache with Caffeine
> -
>
> Key: ARTEMIS-4349
> URL: https://issues.apache.org/jira/browse/ARTEMIS-4349
> Project: ActiveMQ Artemis
> Issue Type: Improvement
>Affects Versions: 2.29.0
>Reporter: Alexey Markevich
>Priority: Major
> Time Spent: 9h 50m
> Remaining Estimate: 0h
>
> based on benchmark https://github.com/ben-manes/caffeine/wiki/Benchmarks
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
[jira] [Resolved] (ARTEMIS-4402) add unit tests to exercise semi-generated LogBundle code more directly
[ https://issues.apache.org/jira/browse/ARTEMIS-4402?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Robbie Gemmell resolved ARTEMIS-4402. - Resolution: Fixed > add unit tests to exercise semi-generated LogBundle code more directly > -- > > Key: ARTEMIS-4402 > URL: https://issues.apache.org/jira/browse/ARTEMIS-4402 > Project: ActiveMQ Artemis > Issue Type: Test > Components: Tests >Reporter: Robbie Gemmell >Assignee: Robbie Gemmell >Priority: Minor > Fix For: 2.31.0 > > > Add unit tests to exercise the semi-generated LogBundle related code more > directly, showing up issues in the generation quicker/more-specifically for > given modules, rather than relying on failures that occur when running much > of the wider more general client+broker test suite (though a lot of which > aren't run in the PR test jobs). -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Commented] (ARTEMIS-4402) add unit tests to exercise semi-generated LogBundle code more directly
[ https://issues.apache.org/jira/browse/ARTEMIS-4402?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17759066#comment-17759066 ] ASF subversion and git services commented on ARTEMIS-4402: -- Commit df410da7aeedcb056d18b9dc625b4a41a2aee500 in activemq-artemis's branch refs/heads/main from Robbie Gemmell [ https://gitbox.apache.org/repos/asf?p=activemq-artemis.git;h=df410da7ae ] ARTEMIS-4402: add some unit tests exercising the semi-generated log/message bundles to make generation problems more quickly noticable > add unit tests to exercise semi-generated LogBundle code more directly > -- > > Key: ARTEMIS-4402 > URL: https://issues.apache.org/jira/browse/ARTEMIS-4402 > Project: ActiveMQ Artemis > Issue Type: Test > Components: Tests >Reporter: Robbie Gemmell >Assignee: Robbie Gemmell >Priority: Minor > Fix For: 2.31.0 > > > Add unit tests to exercise the semi-generated LogBundle related code more > directly, showing up issues in the generation quicker/more-specifically for > given modules, rather than relying on failures that occur when running much > of the wider more general client+broker test suite (though a lot of which > aren't run in the PR test jobs). -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Created] (ARTEMIS-4402) add unit tests to exercise semi-generated LogBundle code more directly
Robbie Gemmell created ARTEMIS-4402: --- Summary: add unit tests to exercise semi-generated LogBundle code more directly Key: ARTEMIS-4402 URL: https://issues.apache.org/jira/browse/ARTEMIS-4402 Project: ActiveMQ Artemis Issue Type: Test Components: Tests Reporter: Robbie Gemmell Assignee: Robbie Gemmell Fix For: 2.31.0 Add unit tests to exercise the semi-generated LogBundle related code more directly, showing up issues in the generation quicker/more-specifically for given modules, rather than relying on failures that occur when running much of the wider more general client+broker test suite (though a lot of which aren't run in the PR test jobs). -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Commented] (ARTEMIS-4391) tests: rework AssertionLoggerHandler
[ https://issues.apache.org/jira/browse/ARTEMIS-4391?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17759053#comment-17759053 ] ASF subversion and git services commented on ARTEMIS-4391: -- Commit 6d4fad7a4c04f1ba96eb3d4fb3d55f6d86787780 in activemq-artemis's branch refs/heads/main from Robbie Gemmell [ https://gitbox.apache.org/repos/asf?p=activemq-artemis.git;h=6d4fad7a4c ] ARTEMIS-4391: fix break in findText from prior changes Only report finding matching log message if all requested entries are present in it, not just the last one provided. Also fix the updated AssertionLoggerHandler usage within AddressFullLoggingTest, ensure it is active across the full period expected messages can happen and doesnt miss early ones. > tests: rework AssertionLoggerHandler > > > Key: ARTEMIS-4391 > URL: https://issues.apache.org/jira/browse/ARTEMIS-4391 > Project: ActiveMQ Artemis > Issue Type: Improvement > Components: Tests >Affects Versions: 2.30.0 >Reporter: Alexey Markevich >Priority: Major > Fix For: 2.31.0 > > Time Spent: 1h 20m > Remaining Estimate: 0h > > AssertionLoggerHandler modified to be easy to use: > * no custom logger configuration required; > * implements Closeable to be poosible to use in try-with-resources statement > * split regular text check and stacktraces -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Moved] (AMQNET-835) Document deserialization policy
[ https://issues.apache.org/jira/browse/AMQNET-835?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Timothy A. Bish moved OPENWIRE-65 to AMQNET-835: Key: AMQNET-835 (was: OPENWIRE-65) Project: ActiveMQ .Net (was: ActiveMQ OpenWire) > Document deserialization policy > --- > > Key: AMQNET-835 > URL: https://issues.apache.org/jira/browse/AMQNET-835 > Project: ActiveMQ .Net > Issue Type: Improvement >Reporter: Arnout Engelen >Priority: Major > > Unrestricted deserialization of untrusted data is dangerous and can lead to > remote code execution attacks. > To be able to safely deserialize untrusted data, the Apache NMS ActiveMQ .Net > client introduced deserialization policy options in version 2.1.0 > ([https://www.mail-archive.com/dev@activemq.apache.org/msg68832.html]). > It would be good to call out in the documentation that if you want to accept > untrusted data, you should use these options. > (I hope this is the correct Jira project to report this to, if not let me > know and I'll re-file it to the correct one :)) -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Resolved] (ARTEMIS-4378) Federation, ignore address policy when using pull consumer connection
[ https://issues.apache.org/jira/browse/ARTEMIS-4378?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Gary Tully resolved ARTEMIS-4378. - Fix Version/s: 2.31.0 Assignee: Gary Tully Resolution: Fixed > Federation, ignore address policy when using pull consumer connection > -- > > Key: ARTEMIS-4378 > URL: https://issues.apache.org/jira/browse/ARTEMIS-4378 > Project: ActiveMQ Artemis > Issue Type: Improvement > Components: Federation >Affects Versions: 2.29.0 >Reporter: Gary Tully >Assignee: Gary Tully >Priority: Minor > Fix For: 2.31.0 > > Time Spent: 50m > Remaining Estimate: 0h > > Using the batching pull consumer from ARTEMIS-4314 is only applicable to > queue Federation but both policies can be configured in the same federation. > If consumer window size of zero is configured any address policy should be > ignored. With address federation there is no local queue to gauge capacity > and messages will just accumate in the upstream. The concept of a pull > consumer for address federation does not make any sense. > This strategy is already adopted for queue Federation configured for > multicast addresses. -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Work logged] (ARTEMIS-4378) Federation, ignore address policy when using pull consumer connection
[ https://issues.apache.org/jira/browse/ARTEMIS-4378?focusedWorklogId=878368=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-878368 ] ASF GitHub Bot logged work on ARTEMIS-4378: --- Author: ASF GitHub Bot Created on: 25/Aug/23 10:15 Start Date: 25/Aug/23 10:15 Worklog Time Spent: 10m Work Description: gtully merged PR #4590: URL: https://github.com/apache/activemq-artemis/pull/4590 Issue Time Tracking --- Worklog Id: (was: 878368) Time Spent: 50m (was: 40m) > Federation, ignore address policy when using pull consumer connection > -- > > Key: ARTEMIS-4378 > URL: https://issues.apache.org/jira/browse/ARTEMIS-4378 > Project: ActiveMQ Artemis > Issue Type: Improvement > Components: Federation >Affects Versions: 2.29.0 >Reporter: Gary Tully >Priority: Minor > Time Spent: 50m > Remaining Estimate: 0h > > Using the batching pull consumer from ARTEMIS-4314 is only applicable to > queue Federation but both policies can be configured in the same federation. > If consumer window size of zero is configured any address policy should be > ignored. With address federation there is no local queue to gauge capacity > and messages will just accumate in the upstream. The concept of a pull > consumer for address federation does not make any sense. > This strategy is already adopted for queue Federation configured for > multicast addresses. -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Commented] (ARTEMIS-4378) Federation, ignore address policy when using pull consumer connection
[ https://issues.apache.org/jira/browse/ARTEMIS-4378?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17758991#comment-17758991 ] ASF subversion and git services commented on ARTEMIS-4378: -- Commit 84c16f1a0d9f930311105e8e430d110ffd8eb058 in activemq-artemis's branch refs/heads/main from Gary Tully [ https://gitbox.apache.org/repos/asf?p=activemq-artemis.git;h=84c16f1a0d ] ARTEMIS-4378 ignore address federation config if connection is configured as pull, consumerWindowSize=0 > Federation, ignore address policy when using pull consumer connection > -- > > Key: ARTEMIS-4378 > URL: https://issues.apache.org/jira/browse/ARTEMIS-4378 > Project: ActiveMQ Artemis > Issue Type: Improvement > Components: Federation >Affects Versions: 2.29.0 >Reporter: Gary Tully >Priority: Minor > Time Spent: 50m > Remaining Estimate: 0h > > Using the batching pull consumer from ARTEMIS-4314 is only applicable to > queue Federation but both policies can be configured in the same federation. > If consumer window size of zero is configured any address policy should be > ignored. With address federation there is no local queue to gauge capacity > and messages will just accumate in the upstream. The concept of a pull > consumer for address federation does not make any sense. > This strategy is already adopted for queue Federation configured for > multicast addresses. -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Created] (AMQ-9303) Main page navigation to Configuration section is broken
Leo Breuss created AMQ-9303: --- Summary: Main page navigation to Configuration section is broken Key: AMQ-9303 URL: https://issues.apache.org/jira/browse/AMQ-9303 Project: ActiveMQ Issue Type: Bug Components: Documentation Reporter: Leo Breuss Broken link in the paragraph "... For details on configuring options, please see refer to the [*Configuration*|https://activemq.apache.org/configuration] section." Page: [https://activemq.apache.org/getting-started#ConfiguringActiveMQ] I find it hard to navigate to the ActiveMQ pages. There is only a very narrow and deeply nested navigation path, ending at said broken link: # [https://activemq.apache.org/] # "ActiveMQ, Find Out More" leads to the features. No direct documentation link available. # button "Read the Docs" --> # "Getting Started" (Note: this is 1 of 20 links, and it's not obvious the full documentation hides behind this navigation option...) # jumping to chapter "[Configuring ActiveMQ"|https://activemq.apache.org/getting-started#ConfiguringActiveMQ], and having to skim through # just to find out the relevant link [*Configuration*|https://activemq.apache.org/configuration] section is broken. -- This message was sent by Atlassian Jira (v8.20.10#820010)
[jira] [Created] (OPENWIRE-65) Document deserialization policy
Arnout Engelen created OPENWIRE-65: -- Summary: Document deserialization policy Key: OPENWIRE-65 URL: https://issues.apache.org/jira/browse/OPENWIRE-65 Project: ActiveMQ OpenWire Issue Type: Improvement Reporter: Arnout Engelen Unrestricted deserialization of untrusted data is dangerous and can lead to remote code execution attacks. To be able to safely deserialize untrusted data, the Apache NMS ActiveMQ .Net client introduced deserialization policy options in version 2.1.0 ([https://www.mail-archive.com/dev@activemq.apache.org/msg68832.html]). It would be good to call out in the documentation that if you want to accept untrusted data, you should use these options. (I hope this is the correct Jira project to report this to, if not let me know and I'll re-file it to the correct one :)) -- This message was sent by Atlassian Jira (v8.20.10#820010)
