[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[ https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14991605#comment-14991605 ] Marcus Olsson commented on CASSANDRA-10242: --- I have a similar issue [here|CASSANDRA-9474], with the difference that the issue is for both the rack and dc validation. I think we should consider implementing the data center validation like it was done for racks here. > Validate rack information on startup > > > Key: CASSANDRA-10242 > URL: https://issues.apache.org/jira/browse/CASSANDRA-10242 > Project: Cassandra > Issue Type: Improvement >Reporter: Jonathan Ellis >Assignee: Carl Yeksigian > Fix For: 3.0.0 rc2, 2.2.4, 2.1.12 > > > Moving to a new rack means that different data should be stored on a node. > We already persist rack information in a system table; we should fail startup > if this doesn't match what the snitch thinks it should be. (Either the > snitch is wrong, and needs to be fixed, or the machine has been moved and > needs to be rebootstrapped.) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[ https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14967788#comment-14967788 ] Aleksey Yeschenko commented on CASSANDRA-10242: --- [~jjordan] Committed as [29ff1f2ac2a3da16f75ce87555df8f6014c8303e|https://github.com/apache/cassandra/commit/29ff1f2ac2a3da16f75ce87555df8f6014c8303e] to 2.1 and merged upwards. > Validate rack information on startup > > > Key: CASSANDRA-10242 > URL: https://issues.apache.org/jira/browse/CASSANDRA-10242 > Project: Cassandra > Issue Type: Improvement >Reporter: Jonathan Ellis >Assignee: Carl Yeksigian > Fix For: 3.0.0 rc2, 2.2.4, 2.1.12 > > > Moving to a new rack means that different data should be stored on a node. > We already persist rack information in a system table; we should fail startup > if this doesn't match what the snitch thinks it should be. (Either the > snitch is wrong, and needs to be fixed, or the machine has been moved and > needs to be rebootstrapped.) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[ https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14967829#comment-14967829 ] Jeremiah Jordan commented on CASSANDRA-10242: - (y) thanks. > Validate rack information on startup > > > Key: CASSANDRA-10242 > URL: https://issues.apache.org/jira/browse/CASSANDRA-10242 > Project: Cassandra > Issue Type: Improvement >Reporter: Jonathan Ellis >Assignee: Carl Yeksigian > Fix For: 3.0.0 rc2, 2.2.4, 2.1.12 > > > Moving to a new rack means that different data should be stored on a node. > We already persist rack information in a system table; we should fail startup > if this doesn't match what the snitch thinks it should be. (Either the > snitch is wrong, and needs to be fixed, or the machine has been moved and > needs to be rebootstrapped.) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[ https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14966103#comment-14966103 ] Jeremiah Jordan commented on CASSANDRA-10242: - The yaml/properties/news entries about this should probably say "decommissioned" not "wiped". Someone is going to read that wrong and lose data. > Validate rack information on startup > > > Key: CASSANDRA-10242 > URL: https://issues.apache.org/jira/browse/CASSANDRA-10242 > Project: Cassandra > Issue Type: Improvement >Reporter: Jonathan Ellis >Assignee: Carl Yeksigian > Fix For: 3.0.0 rc2, 2.2.4, 2.1.12 > > > Moving to a new rack means that different data should be stored on a node. > We already persist rack information in a system table; we should fail startup > if this doesn't match what the snitch thinks it should be. (Either the > snitch is wrong, and needs to be fixed, or the machine has been moved and > needs to be rebootstrapped.) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[ https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14957061#comment-14957061 ] Aleksey Yeschenko commented on CASSANDRA-10242: --- Committed as [711870e3484a7ebec5fe47b2a90bf8f8b3d1e0af|https://github.com/apache/cassandra/commit/711870e3484a7ebec5fe47b2a90bf8f8b3d1e0af] to 2.1 and merged upwards, thank you. > Validate rack information on startup > > > Key: CASSANDRA-10242 > URL: https://issues.apache.org/jira/browse/CASSANDRA-10242 > Project: Cassandra > Issue Type: Improvement >Reporter: Jonathan Ellis >Assignee: Carl Yeksigian > Fix For: 3.0.0 rc2, 2.2.3, 2.1.11 > > > Moving to a new rack means that different data should be stored on a node. > We already persist rack information in a system table; we should fail startup > if this doesn't match what the snitch thinks it should be. (Either the > snitch is wrong, and needs to be fixed, or the machine has been moved and > needs to be rebootstrapped.) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[
https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14954688#comment-14954688
]
Sylvain Lebresne commented on CASSANDRA-10242:
--
Happy to commit this but are we sure we want to take any risk on 2.1 for this?
I'm particularly nervous with that change to
{{conf/cassandra-rackdc.properties}}: are we 100% certain it won't break any
upgrader using {{GossipingPropertyFileSnitch}}?
Also, the branch remove a seemingly unrelated line in the changelog, I suppose
that's an oversight?
> Validate rack information on startup
>
>
> Key: CASSANDRA-10242
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10242
> Project: Cassandra
> Issue Type: Improvement
>Reporter: Jonathan Ellis
>Assignee: Carl Yeksigian
> Fix For: 3.0.0 rc2, 2.2.3, 2.1.11
>
>
> Moving to a new rack means that different data should be stored on a node.
> We already persist rack information in a system table; we should fail startup
> if this doesn't match what the snitch thinks it should be. (Either the
> snitch is wrong, and needs to be fixed, or the machine has been moved and
> needs to be rebootstrapped.)
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[
https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14955153#comment-14955153
]
Carl Yeksigian commented on CASSANDRA-10242:
I replaced the {{conf/cassandra-rackdc.properties}} change with a comment for
people upgrading from {{SimpleSnitch}}, so that it's safe for people who are
already using {{GossipingPropertyFileSnitch}} to upgrade.
Yeah, that was an oversight; I've removed that change.
> Validate rack information on startup
>
>
> Key: CASSANDRA-10242
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10242
> Project: Cassandra
> Issue Type: Improvement
>Reporter: Jonathan Ellis
>Assignee: Carl Yeksigian
> Fix For: 3.0.0 rc2, 2.2.3, 2.1.11
>
>
> Moving to a new rack means that different data should be stored on a node.
> We already persist rack information in a system table; we should fail startup
> if this doesn't match what the snitch thinks it should be. (Either the
> snitch is wrong, and needs to be fixed, or the machine has been moved and
> needs to be rebootstrapped.)
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[ https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14947451#comment-14947451 ] Paulo Motta commented on CASSANDRA-10242: - +1, thanks! Marking as "Ready to Commit". > Validate rack information on startup > > > Key: CASSANDRA-10242 > URL: https://issues.apache.org/jira/browse/CASSANDRA-10242 > Project: Cassandra > Issue Type: Improvement >Reporter: Jonathan Ellis >Assignee: Carl Yeksigian > Fix For: 2.1.x > > > Moving to a new rack means that different data should be stored on a node. > We already persist rack information in a system table; we should fail startup > if this doesn't match what the snitch thinks it should be. (Either the > snitch is wrong, and needs to be fixed, or the machine has been moved and > needs to be rebootstrapped.) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[
https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14947181#comment-14947181
]
Carl Yeksigian commented on CASSANDRA-10242:
Pushed a new [dtest|https://github.com/carlyeks/cassandra-dtest/tree/10242].
I changed the default in {{conf/cassandra-rackdc.properties}} to make the move
from the SimpleSnitch to the GossipingPropertyFileSnitch easier -- these are
the dc and rack which the node has if it was started on SimpleSnitch, so it
won't throw the exception in that case.
> Validate rack information on startup
>
>
> Key: CASSANDRA-10242
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10242
> Project: Cassandra
> Issue Type: Improvement
>Reporter: Jonathan Ellis
>Assignee: Carl Yeksigian
> Fix For: 2.1.x
>
>
> Moving to a new rack means that different data should be stored on a node.
> We already persist rack information in a system table; we should fail startup
> if this doesn't match what the snitch thinks it should be. (Either the
> snitch is wrong, and needs to be fixed, or the machine has been moved and
> needs to be rebootstrapped.)
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[
https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14945762#comment-14945762
]
Paulo Motta commented on CASSANDRA-10242:
-
Patch LGTM overall, could you just add a simple dtest changing the rack of a
single node and check the node will fail to initialize? {{replication_test.py}}
have some example of changing snitch/rack information.
Also, is there any reason you changed the default rack and dc assignments of
{{conf/cassandra-rackdc.properties}} ?
> Validate rack information on startup
>
>
> Key: CASSANDRA-10242
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10242
> Project: Cassandra
> Issue Type: Improvement
>Reporter: Jonathan Ellis
>Assignee: Carl Yeksigian
> Fix For: 2.1.x
>
>
> Moving to a new rack means that different data should be stored on a node.
> We already persist rack information in a system table; we should fail startup
> if this doesn't match what the snitch thinks it should be. (Either the
> snitch is wrong, and needs to be fixed, or the machine has been moved and
> needs to be rebootstrapped.)
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[ https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14877145#comment-14877145 ] Brandon Williams commented on CASSANDRA-10242: -- Some kind of -D override would be nice for the rare situation where you do blow away the data in a dc or something and want to change the racks, but don't want to have to blow away your entire system table and regenerate tokens, etc. > Validate rack information on startup > > > Key: CASSANDRA-10242 > URL: https://issues.apache.org/jira/browse/CASSANDRA-10242 > Project: Cassandra > Issue Type: Improvement >Reporter: Jonathan Ellis >Assignee: Carl Yeksigian > Fix For: 2.1.x > > > Moving to a new rack means that different data should be stored on a node. > We already persist rack information in a system table; we should fail startup > if this doesn't match what the snitch thinks it should be. (Either the > snitch is wrong, and needs to be fixed, or the machine has been moved and > needs to be rebootstrapped.) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[ https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14876872#comment-14876872 ] Jonathan Ellis commented on CASSANDRA-10242: Does it make more sense to put this in SystemKeyspace.checkHealth with the cluster name and partitioner checks? > Validate rack information on startup > > > Key: CASSANDRA-10242 > URL: https://issues.apache.org/jira/browse/CASSANDRA-10242 > Project: Cassandra > Issue Type: Improvement >Reporter: Jonathan Ellis >Assignee: Carl Yeksigian > Fix For: 2.1.x > > > Moving to a new rack means that different data should be stored on a node. > We already persist rack information in a system table; we should fail startup > if this doesn't match what the snitch thinks it should be. (Either the > snitch is wrong, and needs to be fixed, or the machine has been moved and > needs to be rebootstrapped.) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[
https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14804311#comment-14804311
]
Carl Yeksigian commented on CASSANDRA-10242:
I pushed an update which logs the error and exits if it is misconfigured.
I also made it more clear that when switching from {{SimpleSnitch}} ->
{{GossipingPropertyFileSnitch}}, you'll need to make sure that the rack you
specify for existing nodes matches the {{SimpleSnitch}}'s rack1.
> Validate rack information on startup
>
>
> Key: CASSANDRA-10242
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10242
> Project: Cassandra
> Issue Type: Improvement
>Reporter: Jonathan Ellis
>Assignee: Carl Yeksigian
> Fix For: 2.1.x
>
>
> Moving to a new rack means that different data should be stored on a node.
> We already persist rack information in a system table; we should fail startup
> if this doesn't match what the snitch thinks it should be. (Either the
> snitch is wrong, and needs to be fixed, or the machine has been moved and
> needs to be rebootstrapped.)
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-10242) Validate rack information on startup
[ https://issues.apache.org/jira/browse/CASSANDRA-10242?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14790626#comment-14790626 ] Jonathan Ellis commented on CASSANDRA-10242: LGTM in principle, but I don't think AssertionError is appropriate here. Probably ConfigurationException? > Validate rack information on startup > > > Key: CASSANDRA-10242 > URL: https://issues.apache.org/jira/browse/CASSANDRA-10242 > Project: Cassandra > Issue Type: Improvement >Reporter: Jonathan Ellis >Assignee: Carl Yeksigian > Fix For: 2.1.x > > > Moving to a new rack means that different data should be stored on a node. > We already persist rack information in a system table; we should fail startup > if this doesn't match what the snitch thinks it should be. (Either the > snitch is wrong, and needs to be fixed, or the machine has been moved and > needs to be rebootstrapped.) -- This message was sent by Atlassian JIRA (v6.3.4#6332)
