New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Redisson fails to recognize slave that become a master due to master down in Redis Cluster #476
Comments
As I see you have shutdown all nodes for slot range [0-5461], so any objects bounded to this range will be unavailable and all the requests for them fail too. |
I didn't review your log before, sorry. It seems like it's a real bug. Will fix it |
Could you please update to build from "master" and try it due to bug with slots migration. Then please share your log |
Sorry, didn't understand your request. |
Yes |
Fixed. Please try |
This is the log of exactly the same scenario, but with last redisson-2.2.12 (seems like before your fix :) |
I checked again same scenario after your fix with same results, falling requests. Attaching redisson log(after the fix): https://www.dropbox.com/s/eyqllb66qz49w6b/redisson_2.2.12_after_fix.log?dl=0 |
It's strange I don't see "cluster nodes state from" since 20 Apr 2016 14:55:56 from your log. Cluster state update method should try next cluster node to get cluster state. But in your case something goes wrong. |
we also have Connection pool exhausted! error all the time, may be these are related?
|
I'll take a look at this tomorrow |
Important to note that problem exists only if I shutdown redis instances on 10.20.21.113 |
Please try again |
I did same scenario on different environments and the problem always occurred if I'm shutting down redis instances on the middle machine. The scenario: https://www.dropbox.com/s/nei0va82j2hzymo/redis-cluster-middle-host-down.jpg?dl=0 |
+1 for this bug. I get it on stage every time trying to check redis sentinel configuration. |
@ybonda so nothing has changed at all with new fix? |
Not tested yet. Will test it in 10 min |
Will wait for your results |
Yes @mrniko ! Great job man! |
@ybonda At your service! |
@thelamon could you please test it too? |
@mrniko Thanks for the service!Could test it next Monday only ( |
@thelamon OK I'll close it then. If you will have any problems please post a comment |
Hi Niko, seems like problem really exists with redisson driver for cluster.
Here is our configuration:
aa19c3c2a927384b6a4673539cd7dcdb2efa33e8 10.20.21.113:7001 slave d20242223349a4d51b30da40dad09f0dbc9a2cc8 0 1461081575219 1 connected d20242223349a4d51b30da40dad09f0dbc9a2cc8 10.20.21.44:7000 myself,master - 0 0 1 connected 0-5461 ba12c571516b130a9973fdc95ba8cebbfaf5ac7e 10.20.21.113:7000 master - 0 1461081576222 3 connected 5462-10922 4a21be5ffb25369bf4243918c34467103fa9c8a4 10.20.21.59:7001 slave ba12c571516b130a9973fdc95ba8cebbfaf5ac7e 0 1461081577222 3 connected f6b4e5a94409ca397731165acad90f2193154c62 10.20.21.59:7000 master - 0 1461081576422 0 connected 10923-16383 6a7c1e8510d29aed17dc1e0ca7f0e6a7278b4de5 10.20.21.44:7001 slave f6b4e5a94409ca397731165acad90f2193154c62 0 1461081576721 5 connected
I'm shutting down Redis instances on 10.20.21.113:
ba12c571516b130a9973fdc95ba8cebbfaf5ac7e
andaa19c3c2a927384b6a4673539cd7dcdb2efa33e8
When I did this Redis cluster reorganized to following:
aa19c3c2a927384b6a4673539cd7dcdb2efa33e8 10.20.21.113:7001 slave,fail d20242223349a4d51b30da40dad09f0dbc9a2cc8 1461081896304 1461081894700 1 disconnected d20242223349a4d51b30da40dad09f0dbc9a2cc8 10.20.21.44:7000 myself,master - 0 0 1 connected 0-5461 ba12c571516b130a9973fdc95ba8cebbfaf5ac7e 10.20.21.113:7000 master,fail - 1461081896204 1461081895202 3 disconnected 4a21be5ffb25369bf4243918c34467103fa9c8a4 10.20.21.59:7001 master - 0 1461081995791 6 connected 5462-10922 f6b4e5a94409ca397731165acad90f2193154c62 10.20.21.59:7000 master - 0 1461081994787 0 connected 10923-16383 6a7c1e8510d29aed17dc1e0ca7f0e6a7278b4de5 10.20.21.44:7001 slave f6b4e5a94409ca397731165acad90f2193154c62 0 1461081996293 5 connected
As you can see cluster is ok, but requests are failing for some reason.
If I restart tomcat without touching anything on Redis requests are not failing.
Very strange.
This is the link to the redisson log: https://www.dropbox.com/s/pluyzxtxpd0u23z/redisson.log?dl=0
Please assist,
Thank you!
The text was updated successfully, but these errors were encountered: