Add ZFS.md #1694
Add ZFS.md #1694
Conversation
WalkthroughA new file, Changes
Poem
Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media? 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
CodeRabbit Configuration File (
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
Outside diff range, codebase verification and nitpick comments (2)
nodes/zfs.md (2)
25-25: Use a hyphen for the compound adjective "Bare Metal".-1. Bare Metal server with decent amount of RAM (64GB+) +1. Bare-Metal server with decent amount of RAM (64GB+)Tools
LanguageTool
[uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen.
Context: ...ia-bridge/ ``` ::: ## Requirements: 1. Bare Metal server with decent amount of RAM (64GB+...(EN_COMPOUND_ADJECTIVE_INTERNAL)
31-33: Specify the language for shell command code blocks.For better syntax highlighting and readability, specify the language as "sh" for the following code blocks:
- Lines 31-33
- Lines 37-38
- Lines 42-43
- Lines 47-48
- Lines 53-54
- Lines 59-60
- Lines 64-65
- Lines 72-74
- Lines 77-79
- Lines 82-84
- Lines 96-97
Example:
-``` +```sh lsblk --nodeps -o nameAlso applies to: 37-38, 42-43, 47-48, 53-54, 59-60, 64-65, 72-74, 77-79, 82-84, 96-97 </blockquote></details> </blockquote></details> <details> <summary>Review details</summary> **Configuration used: CodeRabbit UI** **Review profile: CHILL** <details> <summary>Commits</summary> Files that changed from the base of the PR and between 95dc48c08af2229039346c9dbb7f25720b0aab54 and 9c2fb73e883a8710626e20548081cf55f90f4388. </details> <details> <summary>Files selected for processing (2)</summary> * nodes/bridge-node.md (1 hunks) * nodes/zfs.md (1 hunks) </details> <details> <summary>Additional context used</summary> <details> <summary>LanguageTool</summary><blockquote> <details> <summary>nodes/zfs.md</summary><blockquote> [uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen. Context: ...ia-bridge/ ``` ::: ## Requirements: 1. Bare Metal server with decent amount of RAM (64GB+... (EN_COMPOUND_ADJECTIVE_INTERNAL) </blockquote></details> <details> <summary>nodes/bridge-node.md</summary><blockquote> [grammar] ~209-~209: The word “setup” is a noun. The verb is spelled with a white space. Context: ...ith ZFS Follow the [tutorial on how to setup your DA node to use on-fly compression ... (NOUN_VERB_CONFUSION) </blockquote></details> </blockquote></details> <details> <summary>Markdownlint</summary><blockquote> <details> <summary>nodes/zfs.md</summary><blockquote> 24-24: Punctuation: ':' Trailing punctuation in heading (MD026, no-trailing-punctuation) --- 28-28: Punctuation: ':' Trailing punctuation in heading (MD026, no-trailing-punctuation) --- 11-11: null Fenced code blocks should have a language specified (MD040, fenced-code-language) --- 18-18: null Fenced code blocks should have a language specified (MD040, fenced-code-language) </blockquote></details> </blockquote></details> </details> <details> <summary>Additional comments not posted (2)</summary><blockquote> <details> <summary>nodes/zfs.md (1)</summary><blockquote> `1-97`: **Documentation looks great!** The documentation provides a clear and comprehensive guide for setting up ZFS compression on a DA node. The step-by-step instructions, command examples for different network environments, and additional notes make it easy for users to follow along and implement ZFS compression on their nodes. Great work on putting together this helpful resource! <details> <summary>Tools</summary> <details> <summary>LanguageTool</summary><blockquote> [uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen. Context: ...ia-bridge/ ``` ::: ## Requirements: 1. Bare Metal server with decent amount of RAM (64GB+... (EN_COMPOUND_ADJECTIVE_INTERNAL) </blockquote></details> <details> <summary>Markdownlint</summary><blockquote> 24-24: Punctuation: ':' Trailing punctuation in heading (MD026, no-trailing-punctuation) --- 28-28: Punctuation: ':' Trailing punctuation in heading (MD026, no-trailing-punctuation) --- 11-11: null Fenced code blocks should have a language specified (MD040, fenced-code-language) --- 18-18: null Fenced code blocks should have a language specified (MD040, fenced-code-language) </blockquote></details> </details> </blockquote></details> <details> <summary>nodes/bridge-node.md (1)</summary><blockquote> `205-209`: **LGTM!** The addition of the optional section for enabling on-the-fly compression using ZFS is a valuable enhancement to the bridge node setup instructions. It provides users with additional options for optimizing their node performance. <details> <summary>Tools</summary> <details> <summary>LanguageTool</summary><blockquote> [grammar] ~209-~209: The word “setup” is a noun. The verb is spelled with a white space. Context: ...ith ZFS Follow the [tutorial on how to setup your DA node to use on-fly compression ... (NOUN_VERB_CONFUSION) </blockquote></details> </details> </blockquote></details> </blockquote></details> </details> <!-- This is an auto-generated comment by CodeRabbit for review status -->
|
hmm @Wondertan I think we'd probably advise not to include this hack in docs? |
|
@jcstein, mentioning that in the docs may be helpful, especially if storage issues are pressing for node runners, as it's ~ a 2x reduction. |
|
@kinrokinro, for how long have you been running the node with compression on? Do you see it running stable? |
For a month and yes, it running stable, our bridge is always in sync and everything looks pretty good to me, that's the reason why I recommend to add this option to the docs. Anyways we always can mark it as "not recommended" if there is any concern from your side. BTW zstd-3 is a light compression, max is zstd-19. |
|
Currently we're running both our bridges on testnet and mainnet on zfs, so our metrics is available in your OTEL |
|
thank you for the context and feedback! I think it is safe to recommend this if @Wondertan approves. And then I will work into the sidebar menu |
I think zstd-3 is optimal as with more zstd-x more time is needed to compress so with bigger load node could be behind sync all the time. |
confirming with @kinrokinro that this uses zstd-3? |
according to the doc that @kinrokinro wrote - yes, it uses zstd-3 and with this compression he achieved 2.05x ratio. According to it level more than zstd-10 leads to crazy time downgrade and ratio isn't so huge to sacrifice with performance. I'm going to try zstd-5 with my mocha bridge node to see if no issues with performance and what compression ratio I will get comparing to zstd-3, probably just after full sync ask to share @kinrokinro and compare with what I got. |
Yes, we use |
|
@kinrokinro do you want to add this to the menu? or think that it is okay as is, linked in the "optional" section? |
This should be optional of course |
|
It would be great to add here that processor is more important than RAM here, tried with 2x Intel Xeon Silver 4108 but it doesn't fly, have to request more powerful. Bridge node just syncs too slow and CPU is always more than 90% while RAM (64GB) about 50%. |
Perhaps it because of more powerful compression you choose |
There was a problem hiding this comment.
Actionable comments posted: 0
Outside diff range, codebase verification and nitpick comments (2)
nodes/zfs.md (2)
24-27: Add a hyphen to "bare metal" and remove the colon.The requirements section provides clear hardware specifications. However, please make the following minor changes:
- Add a hyphen to "bare metal" to make it "bare-metal" when used as a compound adjective.
- Remove the colon at the end of the heading to adhere to the Markdown style guide.
Apply this diff:
-## Requirements: +## Requirements -1. A bare metal server with a substantial amount of RAM (64GB or more) and a modern CPU (latest generation EPYC or Xeon with a clock speed of 2.1GHz or higher and 32 threads or higher is recommended) +1. A bare-metal server with a substantial amount of RAM (64GB or more) and a modern CPU (latest generation EPYC or Xeon with a clock speed of 2.1GHz or higher and 32 threads or higher is recommended)Tools
LanguageTool
[uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen.
Context: ...-bridge/ ``` ::: ## Requirements: 1. A bare metal server with a substantial amount of RAM...(EN_COMPOUND_ADJECTIVE_INTERNAL)
Markdownlint
24-24: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
28-97: Address the static analysis hints and remove the colon from the heading.The guide section is well-structured and provides clear instructions for setting up ZFS compression. The command examples for different network environments are helpful. However, please address the following static analysis hints:
- Remove the colon from the heading to adhere to the Markdown style guide.
- Specify the language for fenced code blocks to improve syntax highlighting.
- Add a comma after "In case of using a snapshot" to improve readability.
Apply this diff:
-## Guide: +## Guide Get your disk name: ```sh @@ -89,7 +89,7 @@ celestia bridge start --node.store /celestia/bridge/.celestia-bridge-arabica-11 ::: :::tip NOTE -It is recommended to sync from scratch. In case of using a snapshot it is important to have your local route to `--data.store` identical to one in a snapshot. +It is recommended to sync from scratch. In case of using a snapshot, it is important to have your local route to `--data.store` identical to one in a snapshot. ::: After completing the steps above, you can begin syncing your DA node.Tools
LanguageTool
[uncategorized] ~45-~45: You might be missing the article “a” here.
Context: ... apt install zfsutils-linuxCreate ZFS pool:sh zpool create $ZFS_POOL_NAME...(AI_EN_LECTOR_MISSING_DETERMINER_A)
[uncategorized] ~89-~89: A comma might be missing here.
Context: ...o sync from scratch. In case of using a snapshot it is important to have your local rout...(AI_EN_LECTOR_MISSING_PUNCTUATION_COMMA)
Markdownlint
28-28: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Files selected for processing (1)
- nodes/zfs.md (1 hunks)
Additional context used
LanguageTool
nodes/zfs.md
[uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen.
Context: ...-bridge/ ``` ::: ## Requirements: 1. A bare metal server with a substantial amount of RAM...(EN_COMPOUND_ADJECTIVE_INTERNAL)
[uncategorized] ~45-~45: You might be missing the article “a” here.
Context: ... apt install zfsutils-linuxCreate ZFS pool:sh zpool create $ZFS_POOL_NAME...(AI_EN_LECTOR_MISSING_DETERMINER_A)
[uncategorized] ~89-~89: A comma might be missing here.
Context: ...o sync from scratch. In case of using a snapshot it is important to have your local rout...(AI_EN_LECTOR_MISSING_PUNCTUATION_COMMA)
Markdownlint
nodes/zfs.md
24-24: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
28-28: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
11-11: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
18-18: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
Additional comments not posted (1)
nodes/zfs.md (1)
1-23: Introduction section looks good!The introduction provides a clear and informative overview of the purpose and benefits of using ZFS compression for a DA node. The note effectively illustrates the storage savings achieved with ZFS compression compared to EXT4 without compression.
Tools
Markdownlint
11-11: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
18-18: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
Will provide here much more info on issues that I got with charts, just still in syncing state. But after switching from 2x Intel Xeon Silver 4108 to 2x Intel Xeon Gold 6130 my node is flying very well, just want to see final result on compression ratio with level what we set (zstd-5). |
not completely, unfortunately. Also adding to decreasing in performance, on the screenshot below you can see that node progresses with 20k blocks every 3 hours now (sadly). Hope more capable I/O drives will solve it. |
based on this will have to recommend not to use this yet, but happy to leave PR open until when it makes sense to add |
Agree, The 4 days are almost passed and node is still in syncing state after switching to new hardware. Less than 1m blocks are left with syncing speed 40k blocks per 12 hours but should say the syncing speed depends on how much data it has so it's faster sometimes. Let's see how it will go after reaching 2m blocks. Will drop here all insights on issues and so on after node will get synced and get confirmed that metrics are being sent well. This testing might be worth to convert into some forum post for user's who will decides to run DA node with zfs later. |
Could you try with zstd-3, not zstd-5? |
Yeah, I left the idea to try syncing with zstd-5 as it requires more processor's resources and maximum that I got compared to zstd-3 is 5% in saving storage. So I will not recommend to try everything higher than zstd-3 or do it on own risk. Now node is on zstd-3 compression from the beginning. |
|
(Moche bridge node) Alright, I've made some investigation previous month and what I've found... Jumping ahead, all things regarding CPU, disks huge usage related ONLY during syncing bridge node, after that it becomes normal and only some really small spikes are happened. ZFS uses certain amount of CPU to process the iops, most of which are compression and checksumming. I used the following hardware setup for my node: Node took about 9 days to get fully synced and here is some hardware resources consumption: Below I would like to list recommendations to start use ZFS.
Below I would like to show I/O wait and disk load with ashift=8 and corresponding disk physical sector size 512-byte. Yeah, NVMe disk with PCIe-4 support could have this :) And with correctly configured zpool ashift properly. The I/O wait in the first case is terrible, max iowait is up to 30% and average - almost 3%, which causes additional load on the CPU and results in the node frequently getting stuck.
It gives me a better compression ratio and seems to reduce the load, as there’s no need to split data into 128KB blocks that could instead be stored in 256KB blocks, resulting in fewer I/O operations. And overall compression. Summarise. After syncing, hardware resource consumption is fine, and the node works well. I know the Celestia team released code that significantly reduced the disk space required for storing data, but considering the 1GB blocks feature that Celestia announced recently, this tool could become relevant again. In this case I plan to use it in the future and recommend it to others. |
|
@mogoll92 @jcstein @kinrokinro Initially, we were stuck grinding through 20,000 blocks in four hours. However, after testing and optimization, we managed to complete the sync in just a couple of hours, last know was around 250,000 blocks every 2 hours or less. One major takeaway was that the high CPU usage we saw was largely due to IOWAIT, which is essentially the system waiting for disk I/O. This can often look like high CPU load on graphs but isn’t actual CPU work. The IOWAIT came from a disk backlog, with tasks sometimes taking 20 to 60 seconds or even longer to process/complete. While upgrading to Gen5 NVMe could have helped, we didn’t want to rely solely on more hardware. Through various tweaks, we reduced the backlog to just 100ms to 4 seconds, leading to much faster sync times. At one point, we tried increasing the recordsize, which improved sync speed even more, but it came at the cost of higher RAM usage and increased disk space. To strike a balance, we decided to keep the default 128K recordsize. It was a long process, but the results speak for themselves! Tested on Gen4 PCIE NVME, DDR4 RAM with Intel XEON
|
I would strongly advise against disabling sync because of data integrity: we are working with the database where data integrity is the main point, if something goes wrong - the database will be corrupted. You should not disable sync no matter what, IMO. |
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Outside diff range and nitpick comments (1)
nodes/bridge-node.md (1)
206-209: Consider rewording "on-fly compression" for clarity.The new section looks good overall. However, consider changing "on-fly compression" to "on-the-fly compression" for better clarity and more common usage.
Apply this diff to improve the wording:
-### Optional: enable on-fly compression with ZFS +### Optional: enable on-the-fly compression with ZFS Follow the -[tutorial on how to set up your DA node to use on-fly compression with ZFS](./zfs.md). +[tutorial on how to set up your DA node to use on-the-fly compression with ZFS](./zfs.md).
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
📒 Files selected for processing (1)
- nodes/bridge-node.md (1 hunks)
🧰 Additional context used
🔇 Additional comments (1)
nodes/bridge-node.md (1)
206-209: LGTM! New section added and past grammar issue fixed.The new section on ZFS compression is a valuable addition to the document. It's clearly marked as optional and provides a link to more detailed instructions. The grammar issue from the past review comment has been successfully addressed.
|
All of these tuning steps (other content is mostly the same as I can see) have been added to the documentation, along with remarks and instructions on how to use the |
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Outside diff range and nitpick comments (5)
nodes/zfs.md (5)
11-16: Minor formatting improvement for code blockThe code block within the tip section could benefit from a minor formatting adjustment for better readability.
Consider applying this change:
-$ zfs get compressratio celestia && du -h /celestia/bridge/.celestia-bridge -NAME PROPERTY VALUE SOURCE -celestia compressratio 2.05x - -2.6T /celestia/bridge/.celestia-bridge +$ zfs get compressratio celestia && du -h /celestia/bridge/.celestia-bridge +NAME PROPERTY VALUE SOURCE +celestia compressratio 2.05x - +2.6T /celestia/bridge/.celestia-bridgeThis aligns the output for better visual clarity.
🧰 Tools
🪛 Markdownlint
11-11: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
24-24: Remove trailing colon from headingThe heading "Requirements:" contains a trailing colon, which is not recommended in Markdown.
Consider changing the heading to:
## RequirementsThis adheres to Markdown best practices and improves consistency.
🧰 Tools
🪛 Markdownlint
24-24: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
28-28: Remove trailing colon from headingThe heading "Guide:" contains a trailing colon, which is not recommended in Markdown.
Consider changing the heading to:
## GuideThis adheres to Markdown best practices and improves consistency.
🧰 Tools
🪛 Markdownlint
28-28: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
11-11: Add language specifiers to code blocksSome code blocks in the document are missing language specifiers. Adding these would improve syntax highlighting and readability.
Consider adding language specifiers to the following code blocks:
- Line 11:
```sh- Line 18:
```sh- Line 32:
```sh- Line 37:
```sh- Line 42:
```sh- Line 47:
```sh- Line 53:
```sh- Line 59:
```sh- Line 64:
```sh- Line 96:
```sh- Line 104:
```sh- Line 114:
```sh- Line 124:
```shThis will ensure consistent syntax highlighting throughout the document.
Also applies to: 18-18, 32-32, 37-37, 42-42, 47-47, 53-53, 59-59, 64-64, 96-96, 104-104, 114-114, 124-124
🧰 Tools
🪛 Markdownlint
11-11: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
101-102: Clarify wording in Auto-Trim sectionThere's a potential typo or unclear wording in the Auto-Trim section.
Consider revising the text to avoid repetition and improve clarity:
-### Disable Auto-Trim -Auto-trim disabling can improve I/O performance, but may lead to increased SSD wear over time. +### Disable Auto-Trim +Disabling auto-trim can improve I/O performance, but may lead to increased SSD wear over time.This change removes the repetition and makes the sentence more straightforward.
🧰 Tools
🪛 LanguageTool
[duplication] ~101-~101: Possible typo: you repeated a word
Context: ...an try the following steps: ### Disable Auto-Trim Auto-trim disabling can improve I/O performance, ...(ENGLISH_WORD_REPEAT_RULE)
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
📒 Files selected for processing (1)
- nodes/zfs.md (1 hunks)
🧰 Additional context used
🪛 LanguageTool
nodes/zfs.md
[uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen.
Context: ...-bridge/ ``` ::: ## Requirements: 1. A bare metal server with a substantial amount of RAM...(EN_COMPOUND_ADJECTIVE_INTERNAL)
[uncategorized] ~89-~89: A comma might be missing here.
Context: ...o sync from scratch. In case of using a snapshot it is important to have your local rout...(AI_EN_LECTOR_MISSING_PUNCTUATION_COMMA)
[duplication] ~101-~101: Possible typo: you repeated a word
Context: ...an try the following steps: ### Disable Auto-Trim Auto-trim disabling can improve I/O performance, ...(ENGLISH_WORD_REPEAT_RULE)
🪛 Markdownlint
nodes/zfs.md
24-24: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
28-28: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
11-11: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
18-18: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
🔇 Additional comments (3)
nodes/zfs.md (3)
30-97: Well-structured guide with clear instructionsThe guide section provides clear, step-by-step instructions for setting up ZFS compression on a DA node. The use of variables, code blocks for different network configurations, and important notes (such as syncing from scratch) are all well-implemented.
The structure and content of this section effectively guide users through the setup process.
🧰 Tools
🪛 LanguageTool
[uncategorized] ~89-~89: A comma might be missing here.
Context: ...o sync from scratch. In case of using a snapshot it is important to have your local rout...(AI_EN_LECTOR_MISSING_PUNCTUATION_COMMA)
99-129: Comprehensive ZFS Fine-Tuning section with clear warningsThe ZFS Fine-Tuning section provides advanced users with valuable optimization options. Each option is well-explained, including potential benefits and risks. The notes on how to revert changes are particularly helpful.
This section effectively balances providing advanced options with necessary cautions, ensuring users can make informed decisions about their ZFS configuration.
🧰 Tools
🪛 LanguageTool
[duplication] ~101-~101: Possible typo: you repeated a word
Context: ...an try the following steps: ### Disable Auto-Trim Auto-trim disabling can improve I/O performance, ...(ENGLISH_WORD_REPEAT_RULE)
1-129: Comprehensive and well-structured guide for ZFS compression on DA nodesThis document successfully addresses the PR objective of introducing documentation for running a Data Availability node with ZFS compression. It provides a clear, step-by-step guide for setup, along with advanced tuning options and important considerations.
Key strengths:
- Clear structure with separate sections for requirements, basic setup, and advanced tuning.
- Effective use of code blocks and tips to highlight important information.
- Balanced approach to presenting advanced options with appropriate warnings.
- Alignment with PR discussion points, such as addressing compression levels (zstd-3) and performance considerations.
While there are minor formatting and wording suggestions mentioned in previous comments, the overall quality and content of the document are excellent. This guide will be a valuable resource for users looking to optimize their DA node performance through ZFS compression.
This document is ready for merging after addressing the minor suggestions in the previous comments.
🧰 Tools
🪛 LanguageTool
[uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen.
Context: ...-bridge/ ``` ::: ## Requirements: 1. A bare metal server with a substantial amount of RAM...(EN_COMPOUND_ADJECTIVE_INTERNAL)
[uncategorized] ~89-~89: A comma might be missing here.
Context: ...o sync from scratch. In case of using a snapshot it is important to have your local rout...(AI_EN_LECTOR_MISSING_PUNCTUATION_COMMA)
[duplication] ~101-~101: Possible typo: you repeated a word
Context: ...an try the following steps: ### Disable Auto-Trim Auto-trim disabling can improve I/O performance, ...(ENGLISH_WORD_REPEAT_RULE)
🪛 Markdownlint
24-24: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
28-28: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
11-11: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
18-18: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
Agree with @kinrokinro that disabling sync may lead to corruption, so need to be careful. Also could you drop some stats, charts, how long has it take, iowait during sync, space taken etc.? |
Could you also add info before creating a pool on ashift and hardware to you doc, please? It's crucial to have the pool with ashift 12 as less values could lead to performance degradation and huge iowait. |
Added, no concerns from my side about this settings (we have |
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Outside diff range and nitpick comments (4)
nodes/zfs.md (4)
9-22: LGTM: Informative compression examplesThe note section effectively demonstrates the benefits of ZFS compression with clear examples. The formatting is consistent and easy to read.
Consider adding a brief explanation of the compression ratio (e.g., "2.05x means the data is compressed to about 48.8% of its original size") to help users interpret the results more easily.
🧰 Tools
🪛 Markdownlint
11-11: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
18-18: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
24-27: LGTM: Clear and specific requirementsThe requirements section provides clear and specific information about the hardware and disk needs for setting up a DA node with ZFS.
Remove the colon from the "Requirements" heading to adhere to Markdown best practices:
-## Requirements: +## Requirements🧰 Tools
🪛 LanguageTool
[uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen.
Context: ...-bridge/ ``` ::: ## Requirements: 1. A bare metal server with a substantial amount of RAM...(EN_COMPOUND_ADJECTIVE_INTERNAL)
🪛 Markdownlint
24-24: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
28-97: LGTM: Comprehensive and well-structured guideThe guide section provides clear, step-by-step instructions for setting up ZFS compression on a DA node. The use of code blocks and network-specific instructions is excellent.
- Remove the colon from the "Guide" heading:
-## Guide: +## Guide
- Consider adding a brief explanation of the
ashiftparameter in the ZFS pool creation command (line 47) to help users understand its purpose and importance.To improve clarity and maintainability, consider using variables for the data store paths in the network-specific commands. For example:
MAINNET_DATA_STORE="/celestia/bridge/.celestia-bridge" MOCHA_DATA_STORE="/celestia/bridge/.celestia-bridge-mocha-4" ARABICA_DATA_STORE="/celestia/bridge/.celestia-bridge-arabica-11" # Then use these variables in the commands, e.g.: celestia bridge start --node.store $MAINNET_DATA_STORE ...This approach would make it easier to update paths in the future if needed.
🧰 Tools
🪛 LanguageTool
[uncategorized] ~89-~89: Possible missing comma found.
Context: ...o sync from scratch. In case of using a snapshot it is important to have your local rout...(AI_HYDRA_LEO_MISSING_COMMA)
🪛 Markdownlint
28-28: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
99-135: LGTM: Comprehensive advanced tuning options with appropriate warningsThe ZFS Fine-Tuning section provides valuable advanced options with clear explanations and appropriate warnings. The structure and use of notes are excellent for guiding users through these more complex configurations.
Fix the typo in the auto-trim note:
-You always can trim maually: `sudo zpool trim $ZFS_POOL_NAME` +You can always trim manually: `sudo zpool trim $ZFS_POOL_NAME`Consider adding a brief explanation of the potential performance impact for each tuning option. For example, for the record size setting:
"Setting a larger record size can improve performance for large, sequential writes but may decrease performance for small, random writes. Choose the appropriate size based on your workload characteristics."
This additional context would help users make informed decisions about which optimizations to apply.
🧰 Tools
🪛 LanguageTool
[duplication] ~101-~101: Possible typo: you repeated a word
Context: ...an try the following steps: ### Disable Auto-Trim Auto-trim disabling can improve I/O performance, ...(ENGLISH_WORD_REPEAT_RULE)
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
📒 Files selected for processing (1)
- nodes/zfs.md (1 hunks)
🧰 Additional context used
🪛 LanguageTool
nodes/zfs.md
[uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen.
Context: ...-bridge/ ``` ::: ## Requirements: 1. A bare metal server with a substantial amount of RAM...(EN_COMPOUND_ADJECTIVE_INTERNAL)
[uncategorized] ~89-~89: Possible missing comma found.
Context: ...o sync from scratch. In case of using a snapshot it is important to have your local rout...(AI_HYDRA_LEO_MISSING_COMMA)
[duplication] ~101-~101: Possible typo: you repeated a word
Context: ...an try the following steps: ### Disable Auto-Trim Auto-trim disabling can improve I/O performance, ...(ENGLISH_WORD_REPEAT_RULE)
🪛 Markdownlint
nodes/zfs.md
24-24: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
28-28: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
11-11: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
18-18: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
🔇 Additional comments (2)
nodes/zfs.md (2)
1-7: LGTM: Clear and informative introductionThe file header and introduction effectively communicate the purpose of the document and the benefits of using ZFS compression for DA nodes. The content is well-structured and provides a good overview for users.
1-135: LGTM: Well-structured and formatted documentThe overall structure and formatting of the document are excellent. The use of Markdown elements, such as headings, code blocks, and tip sections, is consistent and appropriate. The logical flow of information from basic setup to advanced tuning options makes the guide easy to follow.
🧰 Tools
🪛 LanguageTool
[uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen.
Context: ...-bridge/ ``` ::: ## Requirements: 1. A bare metal server with a substantial amount of RAM...(EN_COMPOUND_ADJECTIVE_INTERNAL)
[uncategorized] ~89-~89: Possible missing comma found.
Context: ...o sync from scratch. In case of using a snapshot it is important to have your local rout...(AI_HYDRA_LEO_MISSING_COMMA)
[duplication] ~101-~101: Possible typo: you repeated a word
Context: ...an try the following steps: ### Disable Auto-Trim Auto-trim disabling can improve I/O performance, ...(ENGLISH_WORD_REPEAT_RULE)
🪛 Markdownlint
24-24: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
28-28: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
11-11: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
18-18: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
| ```sh | ||
| ZFS_POOL_NAME="celestia" && ZFS_DATASET_NAME="bridge" | ||
| ``` | ||
|
|
||
| Install ZFS utils: | ||
| ```sh | ||
| sudo apt update && sudo apt install zfsutils-linux | ||
| ``` | ||
|
|
||
| Create ZFS pool: | ||
| ```sh | ||
| zpool create -o ashift=12 $ZFS_POOL_NAME /dev/nvme0n1 | ||
| ``` | ||
|
|
||
| :::tip NOTE | ||
| If you have more than one disk available - you can add them also: | ||
| ```sh | ||
| zpool create -o ashift=12 $ZFS_POOL_NAME /dev/nvme0n1 /dev/nvme1n1 | ||
| ``` | ||
| ::: | ||
|
|
||
| Create dataset: | ||
| ```sh | ||
| zfs create $ZFS_POOL_NAME/$ZFS_DATASET_NAME | ||
| ``` | ||
|
|
||
| Enable compression: | ||
| ```sh | ||
| zfs set compression=zstd-3 $ZFS_POOL_NAME/$ZFS_DATASET_NAME | ||
| ``` | ||
|
|
||
| Set the custom path to the bridge data folder: | ||
|
|
||
| ::: code-group | ||
|
|
||
| ```sh [Mainnet Beta] | ||
| # Add flag --node.store /celestia/bridge/.celestia-bridge to your command, example: | ||
| celestia bridge start --metrics.tls=true --metrics --metrics.endpoint otel.celestia.observer --p2p.metrics --node.store /celestia/bridge/.celestia-bridge | ||
| ``` | ||
|
|
||
| ```sh [Mocha] | ||
| # Add flag --node.store /celestia/bridge/.celestia-bridge-mocha-4 to your command, example: | ||
| celestia bridge start --metrics.tls=true --metrics --metrics.endpoint otel.celestia-mocha.com --p2p.metrics --node.store /celestia/bridge/.celestia-bridge-mocha-4 --p2p.network mocha | ||
| ``` | ||
|
|
||
| ```sh [Arabica] | ||
| # Add flag --node.store /celestia/bridge/.celestia-bridge-arabica-11 to your command, example: | ||
| celestia bridge start --node.store /celestia/bridge/.celestia-bridge-arabica-11 --p2p.network arabica | ||
| ``` | ||
|
|
||
| ::: | ||
|
|
||
| :::tip NOTE | ||
| It is recommended to sync from scratch. In case of using a snapshot it is important to have your local route to `--data.store` identical to one in a snapshot. | ||
| ::: | ||
|
|
||
| After completing the steps above, you can begin syncing your DA node. | ||
|
|
||
| You can check your compression rate with the following command: | ||
| ```sh | ||
| zfs get compressratio $ZFS_POOL_NAME | ||
| ``` | ||
|
|
||
| ## ZFS Fine-Tuning (Advanced) | ||
| If you want to increase your I/O performance and sync speed, you can try the following steps: | ||
| ### Disable Auto-Trim | ||
| Auto-trim disabling can improve I/O performance, but may lead to increased SSD wear over time. | ||
| ```sh | ||
| sudo zpool set autotrim=off $ZFS_POOL_NAME | ||
| ``` | ||
|
|
||
| :::tip NOTE | ||
| You always can trim maually: `sudo zpool trim $ZFS_POOL_NAME` | ||
| ::: | ||
|
|
||
| ### Disable sync | ||
| Disabling boosts write speed, but risks data loss if the system crashes before data is written to disk. | ||
| ```sh | ||
| zfs set sync=disabled $ZFS_POOL_NAME | ||
| ``` | ||
|
|
||
| :::tip NOTE | ||
| You should not keep the `sync` feature disabled permanently; it is useful during the initial DA node sync but can be re-enabled afterward. You can enable `sync` again with: `sudo zfs set sync=enabled $ZFS_POOL_NAME`. | ||
| ::: | ||
|
|
||
| ### Disable prefetch | ||
| Disabling reduces memory usage but can slow down performance for sequential read workloads. | ||
| ```sh | ||
| echo 1 | sudo tee /sys/module/zfs/parameters/zfs_prefetch_disable | ||
| ``` | ||
|
|
||
| :::tip NOTE | ||
| You can always re-enable it: `echo 0 | sudo tee /sys/module/zfs/parameters/zfs_prefetch_disable` | ||
| ::: | ||
|
|
||
| ### Set record size | ||
| Setting `recordsize=256K` defines the maximum block size that ZFS will use when writing data to a dataset. | ||
| ```sh | ||
| zfs set recordsize=256K $ZFS_POOL_NAME/$ZFS_DATASET_NAME | ||
| ``` |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Add a security considerations section
While the document provides comprehensive instructions for setting up and tuning a DA node with ZFS compression, it would benefit from a dedicated section on security considerations.
Consider adding a "Security Considerations" section near the end of the document. This section should cover:
- The potential risks of data loss when disabling sync, emphasizing the importance of having proper backup strategies.
- The importance of securing physical access to the server, especially when using advanced tuning options that may affect data integrity.
- Recommendations for keeping the ZFS software and the underlying operating system up to date with security patches.
- Any specific security best practices related to running a DA node in a ZFS environment.
This addition would help users understand and mitigate potential security risks associated with the setup and tuning options described in the guide.
🧰 Tools
🪛 LanguageTool
[uncategorized] ~25-~25: If this is a compound adjective that modifies the following noun, use a hyphen.
Context: ...-bridge/ ``` ::: ## Requirements: 1. A bare metal server with a substantial amount of RAM...(EN_COMPOUND_ADJECTIVE_INTERNAL)
[uncategorized] ~89-~89: Possible missing comma found.
Context: ...o sync from scratch. In case of using a snapshot it is important to have your local rout...(AI_HYDRA_LEO_MISSING_COMMA)
[duplication] ~101-~101: Possible typo: you repeated a word
Context: ...an try the following steps: ### Disable Auto-Trim Auto-trim disabling can improve I/O performance, ...(ENGLISH_WORD_REPEAT_RULE)
🪛 Markdownlint
24-24: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
28-28: Punctuation: ':'
Trailing punctuation in heading(MD026, no-trailing-punctuation)
11-11: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
18-18: null
Fenced code blocks should have a language specified(MD040, fenced-code-language)
|
I agree that the sync should be enabled after its completed. Here are some stats before optimization
After Optimization
Used space
|
Overview
Add the documentation on how to run a DA node with ZFS compression
Summary by CodeRabbit