Recovery gets stucked during live operations.
SK21
✭
Hello we have noticed that recovery in our vertica-9.0.1.8 gets stucked when we do not take any downtime at
historical completed phase or it keeps starting again and again in loop.
And when we take downtime ie stop all the etls and end users connections we are able to recover the node at once.
Tried select make_ahm_now('true'); but no luck.
0
Answers
THis time it started again in historical phase after almost completing---
Logs keep saying this
2021-04-19 20:19:09.606 CatchUp:7f26d6ffd700-200000000004f1f [Recover] Discarding ROS (200000308ed3131), ROS start=0x1fc6ad6, ROS end=0x1fc6ad6
2021-04-19 20:19:09.606 CatchUp:7f26d6ffd700-200000000004f1f [Recover] Discarding ROS (200000308ed3195), ROS start=0x1fc6ad6, ROS end=0x1fc6ad6
2021-04-19 20:19:09.606 CatchUp:7f26d6ffd700-200000000004f1f [Recover] Discarding ROS (200000308ed31fd), ROS start=0x1fc8f82, ROS end=0x1fc8f82
2021-04-19 20:19:09.607 CatchUp:7f26d6ffd700-200000000004f1f [Recover] Discarding ROS (200000308ed3255), ROS start=0x1fc8f82, ROS end=0x1fc8f82
2021-04-19 20:19:09.607 CatchUp:7f26d6ffd700-200000000004f1f [Recover] Discarding ROS (200000308ed32bd), ROS start=0x1fc8f82, ROS end=0x1fc8f82
2021-04-19 20:19:09.607 CatchUp:7f26d6ffd700-200000000004f1f [Recover] Discarding ROS (200000308ed3327), ROS start=0x1fc8f82, ROS end=0x1fc8f82