Recently I shared several posts about the process to upgrade the GI from 19c to 23ai at ExaCC. My last post summarizes a lot of this, please read it here. But as you know, to use the 23ai you need to be running with OEL 8, and for ExaCC, the upgrade is quite simple. The goal is to reach this, “no updates” and “System up to date”:
With the official release of Oracle 23ai to Exadata on-prem, it is now possible to manually upgrade Grid Infrastructure (GI) from 19c to 23ai. Nowadays the process is simpler than it was in the past, and I already published several examples of how to do this:
So, several examples that you can use as a guide to reach from GI 12.1 to 19c. In this post, I will upgrade from GI 19.23 (19.23.0.0.240416) to GI 23.5 (23.5.0.24.07).
Yes, it is a game-changer. It is for DBAs. It is for DevOps. It is for Enterprises too. And it is not because of new internal services, new ways to access data, or the scalability. But because it changes and improves a lot the way that databases can be refreshed, the way that databases are cloned, how to do CI/CD, and how to deliver databases.
Forget all the slowness and painful process when it is needed to clone production databases over lower environments, or when it is required to clone the dev database to another one. Let’s discover below what can be done with Exascale.
Exascale is built on top of Exadata software. So, all the software features from Exadata are there, the smartscan, the bloom filter, the resource manager, the AI Vector, the JSON, the RDMA, and the RoCE. Even details like the internal services, the MS, CS, and RS continue the same.
On top of that, comes the Exascale software. Several additional services are created to control the communication with the database and deliver the new features. Exascale can be used, deployed, and scaled the way that is needed. It can start, for example, with 300GB until hundreds of terabytes. So, scalability is not an issue.
At Exascale, the usable space is called Vault and the database clusters can share this Vault (imagine that it is the same as ASM diskgroup) to put datafiles redo’s and archivelogs. Going beyond, the storage can be shared (as block devices) by iSCSI to allow plug the Exascale into your network and facilitate the database migrations. When it is OCI, virtual machines can be booted using the Vault as a bootable device.
The communication with databases does not change too much, the Oracle database kernel talks directly with Exascale Vault. So, the first big change, ASM does not exist for 23ai and newer versions. All the redundant processes consuming CPU and memory (by ASM) are gone (imagine all the clusters of ExaCC/ExaCS/Exadata, all of them with their own ASM process). With Exascale they don’t exist anymore because, now, the databases talk directly with Exascale and the Vault. For the 19c database, the ASM is still in place. But at the same Exadata Exascale appliance can have clusters running in 23ai, and others in 19c.
I already wrote about the issue introduced with GI 19.16 in my previous post (click here to read) where (only at Exadata) more space was allocated/reserved by Oracle to guarantee mirror/rebalance. Fortunately, after some months of discussion, they rollbacked the change and released one patch that can be applied at GI 19.19.
The patch was released on 12 of June and it is the number 35285795 and can be only applied at GI 19.19. But to have your space back again there is one important rule: your mirroring needs to be HIGH. This is necessary because the “Smart Rebalance” that allows your disk to be dropped without losing the mirroring. I will write another post just to talk about it.
On 08/March/2023 the Oracle Exadata team released version 23.1.0.0.0 and this include a significant change, OEL 8. I already explained that in my first post that you can read here. In my previous posts, I already described how to patch how to patch storage and switch, and the dom0. In this post, I will discuss how to patch the domU.
I already wrote this previously but is important to understand the upgrade paths that you can do: If you are running the old Exadata with InfiniBand, your dom0 will always be updated until Oracle Linux 7 with UEK5. For domU you can upgrade to the OEL 8. And you can upgrade in any order, first dom0 or domU. If you are running RoCE, your dom0 can run the latest OEL 8 UEK6. The blog post from Oracle made an excellent explanation about the upgrade paths and below you can see the images that are there (I used the image from their post).
On 08/March/2023 the Oracle Exadata team released version 23.1.0.0.0 and this include a significant change, OEL 8. I already explained that in my first post that you can read here. In my second post, I wrote about how to patch storage and switch. In this post, I will discuss how to patch the dom0.
Due to the changes for OEL 8, is important to understand the upgrade paths that you can take. As I wrote in my first post: If you are running the old Exadata with InfiniBand, your dom0 will always be updated until Oracle Linux 7 with UEK5. For domU you can upgrade to the OEL 8. And you can upgrade in any order, first dom0 or domU. If you are running RoCE, your dom0 can run the latest OEL 8 UEK6. The blog post from Oracle made an excellent explanation about the upgrade paths and below you can see the images that are there (I used the image from their post).
So, since the environment that I am patching is Exadata with InfiniBand, my dom0 will be upgraded until the OEL7 running the UEK5. But the Exadata-related software will be upgraded to version 23.1. The domU will be upgraded to OEL8, with UEK6. So, basically will be this (I used the image from the Exadata Team post):
Here, I patched first the dom0 because if I patch it first, all the versions already released for domU will be compatible with him. I am upgrading, so, the dom0 running at 23.1 will be compatible with domU running at a lower version.
On 08/March/2023 the Oracle Exadata team released version 23.1.0.0.0 and this include a significant change, OEL 8. I already explained that in my first post that you can read here. Here I will show how to patch to the 23.1.0.0 version for switch and storage cells.
As you know, I am working with Exadata since 2010 and have already posted about how to upgrade to the 19x version, 18.x version, version 12.x (Portuguese only), and many other details for Oracle Engineered Systems. Fortunately, I had the opportunity to apply the 23.1 version over one environment and will show the details.
Here, my environment is:
Since I am running with dom0/domU, my base machine (from where I will call most of the patches) is the dom0. There, I have ssh passwordless/keyless to all other cells, dbnodes, domU, and switches.
Before you start the patch please check the readme for the patch and identify if you have everything in compliance. Do not start any patch if you meet the requirements. Even from a simple database version, grid, and switch versions. And as well, do not start the patch if your machine has HW errors. So, please read the note Exadata System Software 23.1.0.0.0 Update (32829291) (Doc ID 2772585.1).
On 08/March/2023 the Oracle Exadata team released version 23.1.0.0.0 and this include a significant change, OEL 8. But is not just that, other interesting requirements are there and I will discuss them below. I will show you how to patch to the 23.1 version and some other details as well. In this first part, I will just discuss one interesting point that you need to take care of before you start to patch. And probably is more important than you imagine.
The new version brings some requirements (over what you need to be running) to allow you to patch. For the Grid Infrastructure, you need to run 19.15 or a newer version. You can even run the 21c (21.6 or newer) version if you want. If you want to know how to do that, I already discussed how to upgrade both in previous posts (19c, and 21c).
For databases, the recommendation is the same, 19c or 21c. You can still run older versions (11,g, 12c, and 18) but they are already (or will be soon) under Market Driver Support. You can read the MOS note over that (here), but to be clear (now) only the 19c have premier support available.
And now things became quite interesting because the new 23.1 version is the first running with OEL 8. And if you check the supplemental README for the 23.1 version just the 19c support the database and GI are listed. So, be aware and check the compatibilities.
One important detail for this version is that you can only upgrade to 23.1 if your base Exadata running version is newer or equal to 21.2.10 (basically one year old only). If not, you need to upgrade to (at least) this version before you patch to 23.1. And this will be the same in the one-year future, it will be only possible to upgrade to 24.x if you will be running (at least) 23.1.
If you are running the old Exadata with InfiniBand, your dom0 will always be updated until Oracle Linux 7 with UEK5. For domU you can upgrade to the OEL 8. And you can upgrade in any order, first dom0 or domU. If you are running RoCE, your dom0 can run the latest OEL 8 UEK6. The blog post from Oracle made an excellent explanation about the upgrade paths and below you can see the images that are there (I used the image from their post).
So, as usual, the version includes everything, switches, storage, and database node. And while for switches and storages, the patches are quite normal, for virtualized environments the upgrades paths start to be a little more challenging to plan. I will explain, but (as hinted in the blog post) the upgrade of the Hosts and Guests independently and in any order. And is hard not because of the patch apply itself, but will be to create the plan. Remember the requirements for Oracle Database and GI? So, you can spend a lot of time patching others parts than the Exadata version.
But let’s put pieces together, the small lines written in several places. With this version 23.1, Oracle is telling you that you need to be running at least the Oracle Database 19c to be allowed to have a continuous upgrade for future releases (and possible usage) of Exadata. And whatever the machine version that you use, IB or RoCE network. You can’t anymore use GI older than 19.15, and the databases are enforced, as well, to be this version too. Imagine that you have some kind of incapability between 11g/12c and OEL 8, if you need to open one SR, you need to have/pay for that support, and will not be cheap.
And if you think the upcoming 23c (and that it will be the new LTS version) being in OEL 8 is a requirement. Imagine one year in the future, when the Exadata 24.x version will arrive, do you think that Oracle still supports 11g to the new OEL 9? I don’t think so.
And by the way, IMHO you should be running to 19c. 11G is from 2009, 12.1 from July 2013. So, they are old and out of support for good reasons. I understand the point that they are working on and the legacy applications that maybe you have. But the point is not just to support them, is the case to be possible to continue to upgrade/update your Exadata. Please do not postpone your database upgrades anymore, for the good sake of your Exadata.
Starting with Grid Infrastructure/ASM 19.16 Oracle changed how the REQUIRED_MIRROR_FREE_MB is calculated and the impact is more than expected. Check below examples of the changes, and how this will impact you. This is valid for all GI/ASM starting with 19.16 and only for Exadata/ExaCC.
Please read my new post about this issue.
The REQUIRED_MIRROR_FREE_MB (according to 19c documentation) is:
“amount of space that must be available in a disk group to restore full redundancy after the worst failure that can be tolerated by the disk group without adding additional storage. This requirement ensures that there are sufficient failure groups to restore redundancy”.
And (at Exadata environment until 19.16) is calculated based on the disk redundancy that you have. If you choose the HIGH, the raw size of two disks (the largest in your diskgroup) is reserved; at NORMAL, is the raw size of one disk. At Exadata, it differs from other environments because does not consider the whole failgroup failure and the way that extends are written/spread (more info below and in another post).
But for now, understand that the required size is what you need to reserve (as raw space) at your diskgroup to ensure protection in case of disk failure. And it is directly related to the USABLE_FILE_MB because the space that you can allocate at your diskgroup (USABLE_FILE_MB) comes from (FREE_MB- REQUIRED_MIRROR_FREE_MB)/redundancy factor (3 for HIGH, 2 for NORMAL). So, when you increase the REQUIRED_MIRROR_FREE_MB you reduce the USABLE_FILE_MB. I will explain more later.
Recently I made a tweet about a new project with Oracle Engineered System (X9M) that remembered me about what I made with these systems until now. So, this opened the opportunity to tell my background and history until now working with these systems. Is not a show-off of ego boost post.
