[jira] [Commented] (CASSANDRA-5335) need ability to switch to LCS node by node
[
https://issues.apache.org/jira/browse/CASSANDRA-5335?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13601386#comment-13601386
]
Dean Hiller commented on CASSANDRA-5335:
We went to org.apache.cassandra.db - ColumnFamilies - {our keyspace}- {our
CF} - Attributes - CompactionStrategyClass and then changed SizeTiered to
Leveled or rather the whole string was
org.apache.cassandra.db.compaction.LeveledCompactionStrategy. We hit
enter...nothing happened. We ran a nodetool compact databus5 nreldata and
the files are the same. unfortunately we are in QA with little data and wanted
to set the default size=1 but could not find that in JMX. We can't tell if it
is working or not.
need ability to switch to LCS node by node
--
Key: CASSANDRA-5335
URL: https://issues.apache.org/jira/browse/CASSANDRA-5335
Project: Cassandra
Issue Type: Bug
Affects Versions: 1.2.2
Environment: linux, cassandra 1.2.2
Reporter: Dean Hiller
Labels: compaction
when we switched to LCS, nearly all the servers started rewriting their file
structure. We need a way to switch just one node so we can slowly roll out a
huge change like this. doing the whole cluster brought it to it's needs and
we tried nodetool stop compaction on some nodes but that did not help at all.
We had to really wait overnight for all LCS to finish.
We had also tried nodetool setcompactionthroughput 1 which also had no effect
and performance was still pretty bad. Ideally, none of this would be a
problem if we were just doing one node as we would just pull the node back
out while it was converting to LCS.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (CASSANDRA-5335) need ability to switch to LCS node by node
[ https://issues.apache.org/jira/browse/CASSANDRA-5335?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13600015#comment-13600015 ] Dean Hiller commented on CASSANDRA-5335: that was supposed to read brought it to it's knees not needs need ability to switch to LCS node by node -- Key: CASSANDRA-5335 URL: https://issues.apache.org/jira/browse/CASSANDRA-5335 Project: Cassandra Issue Type: Bug Affects Versions: 1.2.2 Environment: linux, cassandra 1.2.2 Reporter: Dean Hiller Labels: compaction when we switched to LCS, nearly all the servers started rewriting their file structure. We need a way to switch just one node so we can slowly roll out a huge change like this. doing the whole cluster brought it to it's needs and we tried nodetool stop compaction on some nodes but that did not help at all. We had to really wait overnight for all LCS to finish. We had also tried nodetool setcompactionthroughput 1 which also had no effect and performance was still pretty bad. Ideally, none of this would be a problem if we were just doing one node as we would just pull the node back out while it was converting to LCS. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (CASSANDRA-5335) need ability to switch to LCS node by node
Dean Hiller created CASSANDRA-5335: -- Summary: need ability to switch to LCS node by node Key: CASSANDRA-5335 URL: https://issues.apache.org/jira/browse/CASSANDRA-5335 Project: Cassandra Issue Type: Bug Affects Versions: 1.2.2 Environment: linux, cassandra 1.2.2 Reporter: Dean Hiller when we switched to LCS, nearly all the servers started rewriting their file structure. We need a way to switch just one node so we can slowly roll out a huge change like this. doing the whole cluster brought it to it's needs and we tried nodetool stop compaction on some nodes but that did not help at all. We had to really wait overnight for all LCS to finish. We had also tried nodetool setcompactionthroughput 1 which also had no effect and performance was still pretty bad. Ideally, none of this would be a problem if we were just doing one node as we would just pull the node back out while it was converting to LCS. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (CASSANDRA-5333) dynamic snitch only worked for NetworkTopology due to RemoteNode
Dean Hiller created CASSANDRA-5333: -- Summary: dynamic snitch only worked for NetworkTopology due to RemoteNode Key: CASSANDRA-5333 URL: https://issues.apache.org/jira/browse/CASSANDRA-5333 Project: Cassandra Issue Type: Bug Affects Versions: 1.2.2 Environment: redhat linux. we dug into the code though and it appears to be the way it is coded. we were told to file a ticket for this on the email list. Reporter: Dean Hiller We dug into cassandra code and found out SimpleSnitch was not working with DynamicSnitch. Specifically, the code turns a slow node into a remote node and network topology works with that. We also had to move to LOCAL_QUOROM and finally we have dynamic snitch working. To test this, we basically told the NIC using the tc command to slow every packet down by 500ms on node 2 (out of 4 nodes). We finally have a blazingly fast client with node 2 slowed down. If we do QUOROM, the issue comes back and the client is slow again. Same with SimpleSnitch. To reproduce, run a 4 node cluster and use tc to give every packet a slowdown of 500ms. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (CASSANDRA-4580) nodetool can't work from remote node if firewalls are up and only port 7199 is open :(
Dean Hiller created CASSANDRA-4580: -- Summary: nodetool can't work from remote node if firewalls are up and only port 7199 is open :( Key: CASSANDRA-4580 URL: https://issues.apache.org/jira/browse/CASSANDRA-4580 Project: Cassandra Issue Type: Improvement Components: Tools Affects Versions: 1.1.4 Environment: Environment with medium security in which firewalls open specific ports(ie. mostly all production environments from small companies to large). Reporter: Dean Hiller Priority: Minor nodetool uses port 7199 to start with but then JMX opens dynamic ports which can't get through the firewalls. People on the web keep pointing to this ticket which is closed https://issues.apache.org/jira/browse/CASSANDRA-2967 There are definitely a large amount of users wanting this but the ticket appears close and I don't see the replacement ticket(but just close this if you know what it is). This feature is especially needed by operations teams to run the nodetool from their control node. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (CASSANDRA-4580) nodetool can't work from remote node if firewalls are up and only port 7199 is open :(
[ https://issues.apache.org/jira/browse/CASSANDRA-4580?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13443796#comment-13443796 ] Dean Hiller commented on CASSANDRA-4580: I am not sure what you mean by gives me the options??? The ticket was closed and it only seems very related anyways which is why I opened the new one. Are you saying I should try to re-open that ticket? thanks, Dean nodetool can't work from remote node if firewalls are up and only port 7199 is open :( -- Key: CASSANDRA-4580 URL: https://issues.apache.org/jira/browse/CASSANDRA-4580 Project: Cassandra Issue Type: Improvement Components: Tools Affects Versions: 1.1.4 Environment: Environment with medium security in which firewalls open specific ports(ie. mostly all production environments from small companies to large). Reporter: Dean Hiller Priority: Minor nodetool uses port 7199 to start with but then JMX opens dynamic ports which can't get through the firewalls. People on the web keep pointing to this ticket which is closed https://issues.apache.org/jira/browse/CASSANDRA-2967 There are definitely a large amount of users wanting this but the ticket appears close and I don't see the replacement ticket(but just close this if you know what it is). This feature is especially needed by operations teams to run the nodetool from their control node. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Comment Edited] (CASSANDRA-4580) nodetool can't work from remote node if firewalls are up and only port 7199 is open :(
[ https://issues.apache.org/jira/browse/CASSANDRA-4580?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13443796#comment-13443796 ] Dean Hiller edited comment on CASSANDRA-4580 at 8/29/12 3:40 PM: - I am not sure what you mean by gives me the options??? The ticket was closed and it only seems very related anyways which is why I opened the new one. Are you saying I should try to re-open that ticket? OR are you saying that ticket gives me the options to solve the problem myself in which case, I am not sure how they applied to nodetool at allI have nodetool and I have cassandra. The links to documents are too code...am I supposed to modify nodetool then myself? thanks, Dean was (Author: deanhiller): I am not sure what you mean by gives me the options??? The ticket was closed and it only seems very related anyways which is why I opened the new one. Are you saying I should try to re-open that ticket? thanks, Dean nodetool can't work from remote node if firewalls are up and only port 7199 is open :( -- Key: CASSANDRA-4580 URL: https://issues.apache.org/jira/browse/CASSANDRA-4580 Project: Cassandra Issue Type: Improvement Components: Tools Affects Versions: 1.1.4 Environment: Environment with medium security in which firewalls open specific ports(ie. mostly all production environments from small companies to large). Reporter: Dean Hiller Priority: Minor nodetool uses port 7199 to start with but then JMX opens dynamic ports which can't get through the firewalls. People on the web keep pointing to this ticket which is closed https://issues.apache.org/jira/browse/CASSANDRA-2967 There are definitely a large amount of users wanting this but the ticket appears close and I don't see the replacement ticket(but just close this if you know what it is). This feature is especially needed by operations teams to run the nodetool from their control node. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (CASSANDRA-4580) nodetool can't work from remote node if firewalls are up and only port 7199 is open :(
[ https://issues.apache.org/jira/browse/CASSANDRA-4580?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13443800#comment-13443800 ] Dean Hiller commented on CASSANDRA-4580: Also, all those options seems like a ton of work for an operations guy that doesn't really know java.so it really seems like an odd solution for nodetool. nodetool can't work from remote node if firewalls are up and only port 7199 is open :( -- Key: CASSANDRA-4580 URL: https://issues.apache.org/jira/browse/CASSANDRA-4580 Project: Cassandra Issue Type: Improvement Components: Tools Affects Versions: 1.1.4 Environment: Environment with medium security in which firewalls open specific ports(ie. mostly all production environments from small companies to large). Reporter: Dean Hiller Priority: Minor nodetool uses port 7199 to start with but then JMX opens dynamic ports which can't get through the firewalls. People on the web keep pointing to this ticket which is closed https://issues.apache.org/jira/browse/CASSANDRA-2967 There are definitely a large amount of users wanting this but the ticket appears close and I don't see the replacement ticket(but just close this if you know what it is). This feature is especially needed by operations teams to run the nodetool from their control node. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
