1- decides to use blue/green to have no downtime 2- change a parameter in your parameter group for the blue/green 3- needs reboot to apply the change 4- have downtime
Is this because of binlog_format? Facing the same thing How long was your downtime when changing the parameter group, and do you need to leave it a certain time to accumulate the logs?
@@DRim-i2u just done it 30 minutes ago on a 2-node cluster (1 reader / 1 writer, 2 AZs). The param change was very quick, reboot was under a minute. The blue-green setup takes a while, the whole thing was about 90m. Switching took about 20 seconds. We did this under very low load.
It's still useful, but I can't deny the irony of that. Even more that they changed the default on the Aurora 3 parameter group to disable binlogs (or rather rejected the upstream change as it's on by default for MySQL 8 Community), and that's why we now need to reboot a cluster in order to have zero downtime upgrades.
Amazingly I have a AuroraDB v2 cluster using default parameter group which cannot upgrade to v3 at all using all 3 methods. And yes I have posted my question in re:Post which as usual, I have ZERO helpful response at all. What a great user experience.
The following tables show signs that either the datadir directory or the frm file was removed or corrupted. If MySQL encounters this issue, it will write messages to the error logs, which you can check to understand the issue. These tables are considered orphan tables because the respective table data is present in the InnoDB data dictionary but is missing from the MySQL data dictionary.
can we rollback to another aurora2.0 cluster from Aurora 3.0 backup that was upgraded after BGD is completed? or in short, can we we restore aurora3.0 backup to aurora 2.0 instance?
Great question. 🤔 I found this doc that explains the limitations: go.aws/499H3SW. If that doesn't answer your question, reach out to our experts in re:Post: go.aws/aws-repost. 📨 ^RS
A great feature but it's pity that AWS is treating CloudFormation as a second class citizen of the eco-system. It's been more than a year and Blue/Green deployment is not supported.
Sorry to hear about this frustration! I've passed along your sentiment to our team regarding this feature request. You can also keep an eye on our cloudformation-coverage-roadmap via GitHub for updates: go.aws/4fH5pHw. 🤝 ^AD
Hi, Loan! 👋 You can find some info on the upgrade process, here: go.aws/3RiHpjq. 👈 Feel free to reach out to our community of experts on our re:Post platform for more guidance: go.aws/aws-repost. ✍️ ^RF
what if our new green instance is not good enough..we want to roll back to our blue instance..then can we delete green instance????how to come stop blue green deploymemt process????
I'm so sorry for any confusion that may have been caused! 😔 In the description of the video there are some additional links & videos listed for reference that may answer your questions. ⬆️ If that doesn't quite solve the issue, please feel free to reach out to our experts in re:Post: go.aws/aws-repost. They may be able to offer further insight. 📫 ^KR
@@awssupport i want to rectify one question....after we deployed BGD, we find out that green instance not good enough for our application...finally decided to keep our blue instance only..then how to delete green instance..how to delete BGD ???plz rectify
I am terribly sorry for this inconvenience! I've raised your request to our Support team for additional review. 📋 Our community of experts may be able to assist going forward: go.aws/aws-repost. Additionally, we have these self-help options, here: go.aws/get-help. 🛠️☁️ ^KR
Absolutely, we will follow up once we hear back from our Support team about this, should they have additional context to share. ✅ I apologize for the delay in my response. ☁️ ^KR
Hi, this doc may help you troubleshoot: go.aws/49GGn7t. 📁 If you still find yourself needing additonal guidance, please reach out to our collective of engineers via: go.aws/aws-repost. 🔧 ^AR
@@awssupport Thank you for the doc. I was just following your video I wanted to set binlog_format to row as you did but I didn't find binlog_farmat in aurora-mysql8.0 parameter
1- decides to use blue/green to have no downtime
2- change a parameter in your parameter group for the blue/green
3- needs reboot to apply the change
4- have downtime
Is this because of binlog_format?
Facing the same thing
How long was your downtime when changing the parameter group, and do you need to leave it a certain time to accumulate the logs?
had a laugh lmfao
@@DRim-i2u just done it 30 minutes ago on a 2-node cluster (1 reader / 1 writer, 2 AZs). The param change was very quick, reboot was under a minute. The blue-green setup takes a while, the whole thing was about 90m. Switching took about 20 seconds. We did this under very low load.
It's still useful, but I can't deny the irony of that. Even more that they changed the default on the Aurora 3 parameter group to disable binlogs (or rather rejected the upstream change as it's on by default for MySQL 8 Community), and that's why we now need to reboot a cluster in order to have zero downtime upgrades.
@@thiagosestini29 thanks. Ya for us it was also under a minute
Thank you for this video. This helps me upgrading MySQL version for the newest one and it worked. Thanks.
That's fantastic to hear! Glad the video was useful for you. 🙌
Could not find the link for in-place upgrade that you guys mentioned in the video
Amazingly I have a AuroraDB v2 cluster using default parameter group which cannot upgrade to v3 at all using all 3 methods. And yes I have posted my question in re:Post which as usual, I have ZERO helpful response at all. What a great user experience.
The following tables show signs that either the datadir directory or the frm file was removed or corrupted. If MySQL encounters this issue, it will write messages to the error logs, which you can check to understand the issue. These tables are considered orphan tables because the respective table data is present in the InnoDB data dictionary but is missing from the MySQL data dictionary.
after drop the orphan tables, it worked, thanks
can we rollback to another aurora2.0 cluster from Aurora 3.0 backup that was upgraded after BGD is completed? or in short, can we we restore aurora3.0 backup to aurora 2.0 instance?
Great question. 🤔 I found this doc that explains the limitations: go.aws/499H3SW. If that doesn't answer your question, reach out to our experts in re:Post: go.aws/aws-repost. 📨 ^RS
i have upgrade my database from 5.7 to 8.0 but which does it still show RDS Extended Support Enabled?
Hi, our re:Post community alongside our devs may be able to offer further insight into this. Be sure to post your question here: go.aws/3yLaihA. ^CM
these aws relational databases are more costlier than Oracle's
A great feature but it's pity that AWS is treating CloudFormation as a second class citizen of the eco-system. It's been more than a year and Blue/Green deployment is not supported.
Sorry to hear about this frustration! I've passed along your sentiment to our team regarding this feature request. You can also keep an eye on our cloudformation-coverage-roadmap via GitHub for updates: go.aws/4fH5pHw. 🤝 ^AD
@@awssupport Thank you very much :)
You're welcome, we are happy to help! 🙂 ^AR
What if the binlog_format is MIXED?
Hi there! 👋 For the best assistance with your query, kindly post it on our re:Post community of experts, here: go.aws/aws-repost. ^RZ
You can change it. The reason why its needs to be row is because its performing logical replication.
I have a question here, this version upgrade option result in downtime?
Hi, Loan! 👋 You can find some info on the upgrade process, here: go.aws/3RiHpjq. 👈 Feel free to reach out to our community of experts on our re:Post platform for more guidance: go.aws/aws-repost. ✍️ ^RF
what if our new green instance is not good enough..we want to roll back to our blue instance..then can we delete green instance????how to come stop blue green deploymemt process????
I'm so sorry for any confusion that may have been caused! 😔 In the description of the video there are some additional links & videos listed for reference that may answer your questions. ⬆️ If that doesn't quite solve the issue, please feel free to reach out to our experts in re:Post: go.aws/aws-repost. They may be able to offer further insight. 📫 ^KR
@@awssupport i want to rectify one question....after we deployed BGD, we find out that green instance not good enough for our application...finally decided to keep our blue instance only..then how to delete green instance..how to delete BGD ???plz rectify
I am terribly sorry for this inconvenience! I've raised your request to our Support team for additional review. 📋 Our community of experts may be able to assist going forward: go.aws/aws-repost. Additionally, we have these self-help options, here: go.aws/get-help. 🛠️☁️ ^KR
@@awssupport after you get answer , can you plz tell mee..
Absolutely, we will follow up once we hear back from our Support team about this, should they have additional context to share. ✅ I apologize for the delay in my response. ☁️ ^KR
how did you set up your parameter group??
Hi, this doc may help you troubleshoot: go.aws/49GGn7t. 📁 If you still find yourself needing additonal guidance, please reach out to our collective of engineers via: go.aws/aws-repost. 🔧 ^AR
@@awssupport Thank you for the doc. I was just following your video I wanted to set binlog_format to row as you did but I didn't find binlog_farmat in aurora-mysql8.0 parameter
The first video: th-cam.com/video/R5XMny416vo/w-d-xo.html