Forum Discussion
NimbostratusPriority Group Bug
Hi
Yesterday i had an attempt to upgrade the cluster of 2 LTM BIG IP 1600 units. Their current version is 11.4.1 HF3. We wanted to fix the poodle bug of TLS1.x with the recommended HF6. After the upgrade we experienced a problem. The LTM didn't use the logic of the priority groups. Some pools have active members with the high priority and backup members with the low priority. The traffic was distributed between all members of the pool instead of balancing the traffic only between the member wit the highest priority.
The rollback to the previous version solved the problem.
Did someone experience such buggy behavior. I am quite disappointed because last F5 version have a lot of bugs.
Regards
8 Replies
Vitaliy_SavransNacreous
Hi, I have the same issue with the 11.4.1 HF6. Rollback to the previous version is solved priotriy group logic issue. Support team is confirmed that this is a bug.
- Vitaliy_SavransAdditionaly I have periodical CPU utilization of 100%, rollback is solved this also.
Igor_Borovik_72Hi Vitaly Did you receive any solution about that? Regards- Vitaliy_SavransNo, I didn't recieve any solution, yet.
Paul_Barton_829Yes, we upgraded our LTMs from 11.4.1 HF5 to 11.4.1 HF6 and priority activation groups started to ignore priorities. We were forced to downgrade to HF5 which resolved the issue. We opened a ticket with support.
- Igor_Borovik_72
Hi
I got the answer from the F5 support:
As you believe, this is a software issue and we created ID496985 to keep track of this issue. Here is the brief description of the issue.
Issue:Priority group settings are being ignored.
Condition:slow ramp time > 0
Workaround:set slow ramp time to 0Our product development has been working on its fix since it has been reported a few days ago, and it should be available some time next week.
I will be in touch with you as soon as I have more news.
In the mean time, would you please use HF3 or set slow ramp time to 0 for each pool.
- Igor_Borovik_72
Hi
FYI: The hotfix HF7 was released. We installed the hotfix and it solved the issue.
