You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@sentry.apache.org by Arjun Mishra via Review Board <no...@reviews.apache.org> on 2018/01/04 16:40:14 UTC
Re: Review Request 64820: SENTRY-2106: Sentry ahead full snapshot
retry logic
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/
-----------------------------------------------------------
(Updated Jan. 4, 2018, 4:40 p.m.)
Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
Repository: sentry
Description (updated)
-------
After steady state, a full snapshot is triggered by mainly 2 cases
Sentry is "ahead"
Sentry is behind
Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
The solution is to eliminate this check. Sentry can never be ahead of HMS.
Diffs
-----
sentry-hdfs/sentry-hdfs-common/src/main/java/org/apache/sentry/hdfs/ServiceConstants.java a9afb151a
sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
Diff: https://reviews.apache.org/r/64820/diff/3/
Testing
-------
mvn -f sentry-provider/sentry-provider-db/pom.xml test
Thanks,
Arjun Mishra
Re: Review Request 64820: SENTRY-2106: Remove sentry dependency on
HMS table
NOTIFICATION_SEQUENCE when checking if Sentry is out-of-sync with HMS
Posted by Na Li via Review Board <no...@reviews.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/#review195749
-----------------------------------------------------------
Ship it!
Ship It!
- Na Li
On Jan. 16, 2018, 3:50 p.m., Arjun Mishra wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/64820/
> -----------------------------------------------------------
>
> (Updated Jan. 16, 2018, 3:50 p.m.)
>
>
> Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
>
>
> Repository: sentry
>
>
> Description
> -------
>
> After steady state, a full snapshot is triggered by mainly 2 cases
>
> 1. Sentry is "ahead"
> 2. Sentry is behind
> Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
> The fix is to remove this dependency.
>
>
> Diffs
> -----
>
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
> sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
>
>
> Diff: https://reviews.apache.org/r/64820/diff/6/
>
>
> Testing
> -------
>
> mvn -f sentry-provider/sentry-provider-db/pom.xml test
>
>
> Thanks,
>
> Arjun Mishra
>
>
Re: Review Request 64820: SENTRY-2106: Remove sentry dependency on
HMS table
NOTIFICATION_SEQUENCE when checking if Sentry is out-of-sync with HMS
Posted by Vadim Spector via Review Board <no...@reviews.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/#review195750
-----------------------------------------------------------
Ship it!
Ship It!
- Vadim Spector
On Jan. 16, 2018, 3:50 p.m., Arjun Mishra wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/64820/
> -----------------------------------------------------------
>
> (Updated Jan. 16, 2018, 3:50 p.m.)
>
>
> Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
>
>
> Repository: sentry
>
>
> Description
> -------
>
> After steady state, a full snapshot is triggered by mainly 2 cases
>
> 1. Sentry is "ahead"
> 2. Sentry is behind
> Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
> The fix is to remove this dependency.
>
>
> Diffs
> -----
>
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
> sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
>
>
> Diff: https://reviews.apache.org/r/64820/diff/6/
>
>
> Testing
> -------
>
> mvn -f sentry-provider/sentry-provider-db/pom.xml test
>
>
> Thanks,
>
> Arjun Mishra
>
>
Re: Review Request 64820: SENTRY-2106: Remove sentry dependency on
HMS table
NOTIFICATION_SEQUENCE when checking if Sentry is out-of-sync with HMS
Posted by Arjun Mishra via Review Board <no...@reviews.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/
-----------------------------------------------------------
(Updated Jan. 16, 2018, 3:50 p.m.)
Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
Repository: sentry
Description
-------
After steady state, a full snapshot is triggered by mainly 2 cases
1. Sentry is "ahead"
2. Sentry is behind
Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
The fix is to remove this dependency.
Diffs (updated)
-----
sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
Diff: https://reviews.apache.org/r/64820/diff/6/
Changes: https://reviews.apache.org/r/64820/diff/5-6/
Testing
-------
mvn -f sentry-provider/sentry-provider-db/pom.xml test
Thanks,
Arjun Mishra
Re: Review Request 64820: SENTRY-2106: Remove sentry dependency on
HMS table
NOTIFICATION_SEQUENCE when checking if Sentry is out-of-sync with HMS
Posted by Arjun Mishra via Review Board <no...@reviews.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/
-----------------------------------------------------------
(Updated Jan. 16, 2018, 3:50 p.m.)
Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
Summary (updated)
-----------------
SENTRY-2106: Remove sentry dependency on HMS table NOTIFICATION_SEQUENCE when checking if Sentry is out-of-sync with HMS
Repository: sentry
Description
-------
After steady state, a full snapshot is triggered by mainly 2 cases
1. Sentry is "ahead"
2. Sentry is behind
Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
The fix is to remove this dependency.
Diffs
-----
sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
Diff: https://reviews.apache.org/r/64820/diff/5/
Testing
-------
mvn -f sentry-provider/sentry-provider-db/pom.xml test
Thanks,
Arjun Mishra
Re: Review Request 64820: SENTRY-2106: Remove sentry dependency on
HMS table NOTIFICATION_SEQUENCE
Posted by Arjun Mishra via Review Board <no...@reviews.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/
-----------------------------------------------------------
(Updated Jan. 4, 2018, 7:23 p.m.)
Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
Repository: sentry
Description
-------
After steady state, a full snapshot is triggered by mainly 2 cases
1. Sentry is "ahead"
2. Sentry is behind
Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
The fix is to remove this dependency.
Diffs (updated)
-----
sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
Diff: https://reviews.apache.org/r/64820/diff/5/
Changes: https://reviews.apache.org/r/64820/diff/4-5/
Testing
-------
mvn -f sentry-provider/sentry-provider-db/pom.xml test
Thanks,
Arjun Mishra
Re: Review Request 64820: SENTRY-2106: Remove sentry dependency on
HMS table NOTIFICATION_SEQUENCE
Posted by Arjun Mishra via Review Board <no...@reviews.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/
-----------------------------------------------------------
(Updated Jan. 4, 2018, 7:21 p.m.)
Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
Summary (updated)
-----------------
SENTRY-2106: Remove sentry dependency on HMS table NOTIFICATION_SEQUENCE
Repository: sentry
Description (updated)
-------
After steady state, a full snapshot is triggered by mainly 2 cases
1. Sentry is "ahead"
2. Sentry is behind
Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
The fix is to remove this dependency.
Diffs
-----
sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
Diff: https://reviews.apache.org/r/64820/diff/4/
Testing
-------
mvn -f sentry-provider/sentry-provider-db/pom.xml test
Thanks,
Arjun Mishra
Re: Review Request 64820: SENTRY-2106: Eliminate sentry ahead logic
Posted by Na Li via Review Board <no...@reviews.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/#review194770
-----------------------------------------------------------
sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java
Line 263 (original), 263 (patched)
<https://reviews.apache.org/r/64820/#comment273747>
should we get rid of this call for .getCurrentNotificationId?
- Na Li
On Jan. 4, 2018, 5:06 p.m., Arjun Mishra wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/64820/
> -----------------------------------------------------------
>
> (Updated Jan. 4, 2018, 5:06 p.m.)
>
>
> Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
>
>
> Repository: sentry
>
>
> Description
> -------
>
> After steady state, a full snapshot is triggered by mainly 2 cases
>
> 1. Sentry is "ahead"
> 2. Sentry is behind
> Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
> The solution is to eliminate this check. Sentry can never be ahead of HMS.
>
>
> Diffs
> -----
>
> sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
> sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
>
>
> Diff: https://reviews.apache.org/r/64820/diff/4/
>
>
> Testing
> -------
>
> mvn -f sentry-provider/sentry-provider-db/pom.xml test
>
>
> Thanks,
>
> Arjun Mishra
>
>
Re: Review Request 64820: SENTRY-2106: Eliminate sentry ahead logic
Posted by Arjun Mishra via Review Board <no...@reviews.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/
-----------------------------------------------------------
(Updated Jan. 4, 2018, 5:06 p.m.)
Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
Summary (updated)
-----------------
SENTRY-2106: Eliminate sentry ahead logic
Repository: sentry
Description
-------
After steady state, a full snapshot is triggered by mainly 2 cases
1. Sentry is "ahead"
2. Sentry is behind
Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
The solution is to eliminate this check. Sentry can never be ahead of HMS.
Diffs
-----
sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
Diff: https://reviews.apache.org/r/64820/diff/4/
Testing
-------
mvn -f sentry-provider/sentry-provider-db/pom.xml test
Thanks,
Arjun Mishra
Re: Review Request 64820: SENTRY-2106: Sentry ahead full snapshot
retry logic
Posted by Arjun Mishra via Review Board <no...@reviews.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/
-----------------------------------------------------------
(Updated Jan. 4, 2018, 5:05 p.m.)
Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
Changes
-------
Removing the sentry "ahead" logic
Repository: sentry
Description
-------
After steady state, a full snapshot is triggered by mainly 2 cases
1. Sentry is "ahead"
2. Sentry is behind
Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
The solution is to eliminate this check. Sentry can never be ahead of HMS.
Diffs (updated)
-----
sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
Diff: https://reviews.apache.org/r/64820/diff/4/
Changes: https://reviews.apache.org/r/64820/diff/3-4/
Testing
-------
mvn -f sentry-provider/sentry-provider-db/pom.xml test
Thanks,
Arjun Mishra
Re: Review Request 64820: SENTRY-2106: Sentry ahead full snapshot
retry logic
Posted by Arjun Mishra via Review Board <no...@reviews.apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/64820/
-----------------------------------------------------------
(Updated Jan. 4, 2018, 4:43 p.m.)
Review request for sentry, Alexander Kolbasov, Brian Towles, kalyan kumar kalvagadda, Na Li, Sergio Pena, and Vadim Spector.
Repository: sentry
Description (updated)
-------
After steady state, a full snapshot is triggered by mainly 2 cases
1. Sentry is "ahead"
2. Sentry is behind
Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are not in sync. As a result of this unnecessary full snapshots can be triggered.
The solution is to eliminate this check. Sentry can never be ahead of HMS.
Diffs
-----
sentry-hdfs/sentry-hdfs-common/src/main/java/org/apache/sentry/hdfs/ServiceConstants.java a9afb151a
sentry-provider/sentry-provider-db/src/main/java/org/apache/sentry/service/thrift/HMSFollower.java aa1b6a31c
sentry-provider/sentry-provider-db/src/test/java/org/apache/sentry/service/thrift/TestHMSFollower.java edde886a7
Diff: https://reviews.apache.org/r/64820/diff/3/
Testing
-------
mvn -f sentry-provider/sentry-provider-db/pom.xml test
Thanks,
Arjun Mishra