Seagate / halon

High availability solution
Apache License 2.0
1 stars 0 forks source link

Problem: CI tests fail #1591

Closed vvv closed 4 years ago

vvv commented 4 years ago

E.g., “ST singlenode” job fails with the following error message in syslog:

Sep 23 11:19:45.136605 halon50778-s-cmu mero-server[3083]: mero[03083]:  9780  FATAL  [lib/assert.c:48:m0_panic]  panic: (({ typeof (msg->hm_data.u.hed_nvec.hmnv_ignore_same_state) __x = (msg->hm_data.u.hed_nvec.hmnv_ignore_same_state); ((__x) == (0) || ((__x) == (1))); })) at m0_ha_msg_accept() (ha/note.c:198)  [git: 3c89e4b] /var/mero/m0d-0x7200000000000001:0x1b/m0trace.3083
Sep 23 11:19:45.136879 halon50778-s-cmu mero-server[3083]: Mero panic: (({ typeof (msg->hm_data.u.hed_nvec.hmnv_ignore_same_state) __x = (msg->hm_data.u.hed_nvec.hmnv_ignore_same_state); ((__x) == (0) || ((__x) == (1))); })) at m0_ha_msg_accept() ha/note.c:198 (errno: 0) (last failed: none) [git: 3c89e4b] pid: 3083  /var/mero/m0d-0x7200000000000001:0x1b/m0trace.3083

Apparently halond sends garbage in m0_ha_msg_nvec::hmnv_ignore_same_state field.

Solution: prior to sending m0_ha_msg_nvec to Mero, set its hmnv_ignore_same_state field to 0 (false).

hmnv_ignore_same_state field has been added by Mero commit mero/mero@08efd9f25fbc03f7e742957497515d68d3fc61f2 (c/18379).

andriytk commented 4 years ago

merged

vvv commented 4 years ago

changed the description

vvv commented 4 years ago

changed title from {-HAState.hsc (notify): Set hmnv_ignore_same_state-} to {+Problem: CI tests fail+}

andriytk commented 4 years ago

enabled an automatic merge when the pipeline for 16022535351234bb3c8042ba97931eb22d942682 succeeds

vvv commented 4 years ago

cc @max-seagate.medved