[
https://issues.apache.org/jira/browse/CASSANDRA-18438?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17742299#comment-17742299
]
Jackson Fleming edited comment on CASSANDRA-18438 at 7/12/23 7:37 AM:
----------------------------------------------------------------------
Tested the GCP snitch, looks good on my end
!image-2023-07-12-17-28-24-723.png!
The following log line was present on startup
{code:java}
INFO [main] 2023-07-12 07:23:09,575 AbstractCloudMetadataServiceSnitch.java:59
- org.apache.cassandra.locator.GoogleCloudSnitch using datacenter: us-central1,
rack: a, connector:
org.apache.cassandra.locator.AbstractCloudMetadataServiceConnector$DefaultCloudMetadataServiceConnector{metadata_url=http://metadata.google.internal,metadata_request_timeout=30000},
properties: SnitchProperties{properties={rack=rack1, dc=dc1}} {code}
The following test steps were performed:
1 - Sanity testRan with SimpleSnitch to make sure Cassandra is built and setup
correctly.
{code:java}
~/apache-cassandra-5.0-SNAPSHOT/bin$ ./nodetool status
Datacenter: datacenter1
=======================
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
-- Address Load Tokens Owns (effective) Host ID
Rack
UN 127.0.0.1 116.43 KiB 16 100.0%
3b90fbed-fff1-4499-84b1-28cfe63cd93e rack1
{code}
So we know it's built correctly and working.
2 - Next update the snitch name to GoogleCloudSnitch in cassandra.yml. Clear
the data directories with rm -rf data/*, start up Cassandra. Run nodetool status
{code:java}
~/apache-cassandra-5.0-SNAPSHOT/bin$ ./nodetool status
Datacenter: us-central1
=======================
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
-- Address Load Tokens Owns (effective) Host ID
Rack
UN 127.0.0.1 116.4 KiB 16 100.0%
ed234a02-4af3-433c-950f-cbec23c87fb6 a {code}
I have attached full startup logs as well, incase they are useful.
[https://pastebin.com/LUM42X3W] (this Jira won't let me upload a log file, if
pastebin is not appropriate just let me know how I can provide them).
was (Author: JIRAUSER294685):
Tested the GCP snitch, looks good on my end
!image-2023-07-12-17-28-24-723.png!
The following log line was present on startup
{code:java}
INFO [main] 2023-07-12 07:23:09,575 AbstractCloudMetadataServiceSnitch.java:59
- org.apache.cassandra.locator.GoogleCloudSnitch using datacenter: us-central1,
rack: a, connector:
org.apache.cassandra.locator.AbstractCloudMetadataServiceConnector$DefaultCloudMetadataServiceConnector{metadata_url=http://metadata.google.internal,metadata_request_timeout=30000},
properties: SnitchProperties{properties={rack=rack1, dc=dc1}} {code}
The following test steps were performed:
1 - Sanity testRan with SimpleSnitch to make sure Cassandra is built and setup
correctly.
~/apache-cassandra-5.0-SNAPSHOT/bin$ ./nodetool status
Datacenter: datacenter1
=======================
Status=Up/Down
|/ State=Normal/Leaving/Joining/Moving
– Address Load Tokens Owns (effective) Host ID Rack
UN 127.0.0.1 116.43 KiB 16 100.0% 3b90fbed-fff1-4499-84b1-28cfe63cd93e rack1
So we know that’s working.
2 - Next update the snitch name to GoogleCloudSnitch in cassandra.yml. Clear
the data directories with rm -rf data/*, start up Cassandra. Run nodetool status
~/apache-cassandra-5.0-SNAPSHOT/bin$ ./nodetool status
Datacenter: us-central1
=======================
Status=Up/Down|
|/ State=Normal/Leaving/Joining/Moving
– Address Load Tokens Owns (effective) Host ID Rack
UN 127.0.0.1 116.4 KiB 16 100.0% ed234a02-4af3-433c-950f-cbec23c87fb6 a|
I have attached full startup logs as well, incase they are useful.
[https://pastebin.com/LUM42X3W] (this Jira won't let me upload a log file, if
pastebin is not appropriate just let me know how I can provide them)
> Refactor cloud snitches to get rid of duplicate code
> ----------------------------------------------------
>
> Key: CASSANDRA-18438
> URL: https://issues.apache.org/jira/browse/CASSANDRA-18438
> Project: Cassandra
> Issue Type: Improvement
> Components: Local/Other
> Reporter: Maxwell Guo
> Assignee: Stefan Miklosovic
> Priority: Normal
> Fix For: 5.x
>
> Attachments: alicloud-snitch.png, image-2023-07-12-17-28-24-723.png,
> image-2023-07-12-17-30-53-158.png, image-2023-07-12-17-31-37-447.png
>
> Time Spent: 5h 50m
> Remaining Estimate: 0h
>
> Nowadays we have got about four public cloud platform snitchs ： EC2 snitch
> for aws, google cloud snitch for google cloud, alibaba cloud snitch for
> alibaba cloud and multi region snitch for ec2. And the common place for the
> first three is that we just need to query the zone center to get the ec2 /
> ecs id , so I think we can refactor the code , and if some new public cloud
> platform want to add one more snitch for himself, there is no need to pull a
> pr for him and configure some options in yaml is enough .
> Besides it would be even better that we may reuse the multic region snitch
> for ec2 for other public cloud platform.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
