-
Notifications
You must be signed in to change notification settings - Fork 177
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
explict wsrep_sst_donor fails when node doesn't exist with un-intuitive error message #108
Comments
https://mariadb.atlassian.net/browse/MDEV-6699 of donor. | Variable_name | Value | db-prod-iin-galera3 also 5.5.39-MariaDB-1~precise-wsrep-log |
Faking -EHOSTUNREACH, -EAGAIN and -EHOSTDOWN wasn't intuitative when debugging. Closes codership#108
+1 It took me a while to debug this error message. I had defined wsrep-sst-donor by IP address, and after seeing that error message I did all sort of network debugging until I noticed that the donor must be specified by node name as defined in the wsrep_node_name parameter. |
#109 contains a fix. I wrote it to be minimal however if something different is required just ask. |
Can I +1 to this being fixed sooner rather than later ( If it helps I can reference a mariadb.com support ticket ). |
Merge with 3.12
I updating to mariadb-galera-5.5.39 and came across the following error.
Sep 5 09:58:59 db-prod-iin-galera3 mysqld: 140905 9:58:59 [Note] WSREP: Prepared IST receiver, listening at: tcp://10.1.2.113:4568
Sep 5 09:58:59 db-prod-iin-galera3 mysqld: 140905 9:58:59 [ERROR] WSREP: Requesting state transfer failed: -113(No route to host)
Sep 5 09:58:59 db-prod-iin-galera3 mysqld: 140905 9:58:59 [ERROR] WSREP: State transfer request failed unrecoverably: 113 (No route to host). Most likely it is due to inability to communicate with the cluster primary component. Restart required.
Sep 5 09:58:59 db-prod-iin-galera3 mysqld: 140905 9:58:59 [Note] WSREP: Closing send monitor...
After much examination and strace-ing of mysqld on this instance and the donor neither side reported EHOSTUNREACH (wsrep_sst_donor='db-prod--iin-rep1' was specified).
bug 1: miss leading error messages
returning EHOSTUNREACH in ./gcs/src/gcs_group.cpp:group_find_node_by_name - can an explicit gu_info/gu_error message be called within that function that describes the message that node db-prod--iin-rep1 could not be found.
(And for this Arjen ranted http://openquery.com.au/blog/tracing-problem-finding-sloppy-code )
bug 1a:
http://galeracluster.com/documentation-webpages/mysqlwsrepoptions.html#wsrep-sst-donor
-11(Resource temporarily unavailable) (does this still apply?). -EHOSTDOWN, gcs_group_act_conf (-ENOMEM) are also bizarrely cryptic.
The text was updated successfully, but these errors were encountered: