Bluestore fried chicken + wingstop ranch = taste of heaven
December 21, 2024 at 11:48 PM
Bluestore fried chicken + wingstop ranch = taste of heaven
Conventional wisdom says standing on the shoulders of giants—leveraging battle-tested technologies. However, it's worth re-evaluating that decision when prior art becomes a substantial blocker. Sometimes, you need a clean-slate approach. This paper shows a good example: why/how Ceph built BlueStore.
March 30, 2025 at 7:20 PM
Conventional wisdom says standing on the shoulders of giants—leveraging battle-tested technologies. However, it's worth re-evaluating that decision when prior art becomes a substantial blocker. Sometimes, you need a clean-slate approach. This paper shows a good example: why/how Ceph built BlueStore.
i miss bluestore.
November 29, 2024 at 8:35 PM
i miss bluestore.
Your posts here prove you know that analogy doesn't work. You are currently selling food at Nazi-Mart, and BlueStore. You just don't make as much money at BlueStore as you do at Nazi-Mart.
May 26, 2026 at 1:54 PM
Your posts here prove you know that analogy doesn't work. You are currently selling food at Nazi-Mart, and BlueStore. You just don't make as much money at BlueStore as you do at Nazi-Mart.
Vai no play store procura o app Bluestore social e clica em atualiza
April 26, 2024 at 10:29 AM
Vai no play store procura o app Bluestore social e clica em atualiza
I enjoy reading the story of how Ceph initially tried to build the storage backend on top of general file systems (e.g., ext4 or XFS), struggled with performance and complexity, and eventually decided to develop a new storage backend, BlueStore, directly built on top of raw storage devices.
March 30, 2025 at 7:20 PM
I enjoy reading the story of how Ceph initially tried to build the storage backend on top of general file systems (e.g., ext4 or XFS), struggled with performance and complexity, and eventually decided to develop a new storage backend, BlueStore, directly built on top of raw storage devices.
I have an array of them on a pci m.2 riser.
I'm about to start hacking on a ceph BlueStore replacement and I want the optane ddr + optane u.2 to tune the entire L1 > L2 > L3 > DDR4 > Optane DDR4 > optane u.2 > BlueField DPU > SFP datapath.
A girl has aquisition problems.
I'm about to start hacking on a ceph BlueStore replacement and I want the optane ddr + optane u.2 to tune the entire L1 > L2 > L3 > DDR4 > Optane DDR4 > optane u.2 > BlueField DPU > SFP datapath.
A girl has aquisition problems.
August 1, 2026 at 12:46 AM
I have an array of them on a pci m.2 riser.
I'm about to start hacking on a ceph BlueStore replacement and I want the optane ddr + optane u.2 to tune the entire L1 > L2 > L3 > DDR4 > Optane DDR4 > optane u.2 > BlueField DPU > SFP datapath.
A girl has aquisition problems.
I'm about to start hacking on a ceph BlueStore replacement and I want the optane ddr + optane u.2 to tune the entire L1 > L2 > L3 > DDR4 > Optane DDR4 > optane u.2 > BlueField DPU > SFP datapath.
A girl has aquisition problems.
To address these issues, Ceph built BlueStore. It implements a special-purpose interface in user space, avoiding most of the inefficiency and complexity of general-purpose FS. Since it has full control of the I/O stack, it enables new features and the adoption of emerging new hardware.
March 30, 2025 at 7:20 PM
To address these issues, Ceph built BlueStore. It implements a special-purpose interface in user space, avoiding most of the inefficiency and complexity of general-purpose FS. Since it has full control of the I/O stack, it enables new features and the adoption of emerging new hardware.
Of course, it faces new challenges when operating raw storage devices. For example, BlueStore can no longer rely on the OS page cache, which can lead to increased CPU and memory usage. Ceph has been adding auto-tuning features over the years – I'm looking forward to seeing more work in this area.
March 30, 2025 at 7:20 PM
Of course, it faces new challenges when operating raw storage devices. For example, BlueStore can no longer rely on the OS page cache, which can lead to increased CPU and memory usage. Ceph has been adding auto-tuning features over the years – I'm looking forward to seeing more work in this area.
🐙 #Cephalocon Spotlight
BlueStore: Faster + Smaller = Better – Adam Kupczyk & Jaya Prakash Madaka, IBM
Optimizations that cut TCO and boost storage efficiency.
Read more here: t.ly/cephalocon-s...
#Ceph #OpenSource
BlueStore: Faster + Smaller = Better – Adam Kupczyk & Jaya Prakash Madaka, IBM
Optimizations that cut TCO and boost storage efficiency.
Read more here: t.ly/cephalocon-s...
#Ceph #OpenSource
September 30, 2025 at 1:15 PM
🐙 #Cephalocon Spotlight
BlueStore: Faster + Smaller = Better – Adam Kupczyk & Jaya Prakash Madaka, IBM
Optimizations that cut TCO and boost storage efficiency.
Read more here: t.ly/cephalocon-s...
#Ceph #OpenSource
BlueStore: Faster + Smaller = Better – Adam Kupczyk & Jaya Prakash Madaka, IBM
Optimizations that cut TCO and boost storage efficiency.
Read more here: t.ly/cephalocon-s...
#Ceph #OpenSource
slamming 25gb of writes to the bucket as fast as they network allowed resulted in this alert in the dashboard:
` BLUESTORE_SLOW_OP_ALERT: 4 OSD(s) experiencing slow operations in BlueStore `
hmmm.
` BLUESTORE_SLOW_OP_ALERT: 4 OSD(s) experiencing slow operations in BlueStore `
hmmm.
March 22, 2025 at 1:57 PM
slamming 25gb of writes to the bucket as fast as they network allowed resulted in this alert in the dashboard:
` BLUESTORE_SLOW_OP_ALERT: 4 OSD(s) experiencing slow operations in BlueStore `
hmmm.
` BLUESTORE_SLOW_OP_ALERT: 4 OSD(s) experiencing slow operations in BlueStore `
hmmm.
Ceph v18.2.8 (Reef) is out! 🚀
The final expected Reef backport release brings:
🔐 Security fixes (CephFS, mgr alerts)
⚙️ Stability + correctness improvements across RGW, CephFS, BlueStore
Read more: t.ly/Reefv1828
#Ceph #OpenSource #Storage
The final expected Reef backport release brings:
🔐 Security fixes (CephFS, mgr alerts)
⚙️ Stability + correctness improvements across RGW, CephFS, BlueStore
Read more: t.ly/Reefv1828
#Ceph #OpenSource #Storage
March 23, 2026 at 4:50 PM
Ceph v18.2.8 (Reef) is out! 🚀
The final expected Reef backport release brings:
🔐 Security fixes (CephFS, mgr alerts)
⚙️ Stability + correctness improvements across RGW, CephFS, BlueStore
Read more: t.ly/Reefv1828
#Ceph #OpenSource #Storage
The final expected Reef backport release brings:
🔐 Security fixes (CephFS, mgr alerts)
⚙️ Stability + correctness improvements across RGW, CephFS, BlueStore
Read more: t.ly/Reefv1828
#Ceph #OpenSource #Storage
🐙 #Ceph v19.2.4 Squid is out!
This release delivers important fixes for:
✅ #BlueStore & #BlueFS
✅ #RGW notifications and multipart uploads
✅ #CephFS & MDS diagnostics
✅ And more...
Read the full breakdown here: t.ly/SquidRelease4
#OpenSource #Storage
This release delivers important fixes for:
✅ #BlueStore & #BlueFS
✅ #RGW notifications and multipart uploads
✅ #CephFS & MDS diagnostics
✅ And more...
Read the full breakdown here: t.ly/SquidRelease4
#OpenSource #Storage
June 1, 2026 at 2:56 PM
🐙 #Ceph v19.2.4 Squid is out!
This release delivers important fixes for:
✅ #BlueStore & #BlueFS
✅ #RGW notifications and multipart uploads
✅ #CephFS & MDS diagnostics
✅ And more...
Read the full breakdown here: t.ly/SquidRelease4
#OpenSource #Storage
This release delivers important fixes for:
✅ #BlueStore & #BlueFS
✅ #RGW notifications and multipart uploads
✅ #CephFS & MDS diagnostics
✅ And more...
Read the full breakdown here: t.ly/SquidRelease4
#OpenSource #Storage
View the tracks for KCD x Ceph x OpenInfra Day Korea 2026!
7 #Ceph sessions: multi-cluster orchestration, #BlueStore tuning, #CephFS for AI, 4PB Rook-Ceph at scale, and more.
📅 Sept 1, Seoul
🔗 : t.ly/kcd-ceph-ope...
#CephCommunity #Kubernetes #OpenInfra #OpenSource
7 #Ceph sessions: multi-cluster orchestration, #BlueStore tuning, #CephFS for AI, 4PB Rook-Ceph at scale, and more.
📅 Sept 1, Seoul
🔗 : t.ly/kcd-ceph-ope...
#CephCommunity #Kubernetes #OpenInfra #OpenSource
August 13, 2026 at 12:51 PM
View the tracks for KCD x Ceph x OpenInfra Day Korea 2026!
7 #Ceph sessions: multi-cluster orchestration, #BlueStore tuning, #CephFS for AI, 4PB Rook-Ceph at scale, and more.
📅 Sept 1, Seoul
🔗 : t.ly/kcd-ceph-ope...
#CephCommunity #Kubernetes #OpenInfra #OpenSource
7 #Ceph sessions: multi-cluster orchestration, #BlueStore tuning, #CephFS for AI, 4PB Rook-Ceph at scale, and more.
📅 Sept 1, Seoul
🔗 : t.ly/kcd-ceph-ope...
#CephCommunity #Kubernetes #OpenInfra #OpenSource
Upgrade your business connectivity with The Blue Store, your ultimate destination for modern networking solutions. Discover the power of technology that works as hard as you do every day. #BlueStore #NetworkingSolutions #TechUpgrade
https://empro.com/empro-tech-blog/the-best-cloud-hosting-services-
https://empro.com/empro-tech-blog/the-best-cloud-hosting-services-
July 2, 2026 at 12:10 PM
Upgrade your business connectivity with The Blue Store, your ultimate destination for modern networking solutions. Discover the power of technology that works as hard as you do every day. #BlueStore #NetworkingSolutions #TechUpgrade
https://empro.com/empro-tech-blog/the-best-cloud-hosting-services-
https://empro.com/empro-tech-blog/the-best-cloud-hosting-services-
I sized every Ceph database at 4% and reserved 2,887 GiB of NVMe. Nineteen of them use 87.
The docs weren't wrong. I'd treated the conservative number like a measurement of my workload.
Then the 500x benchmark failed the same test. I'd measured a sick cluster, then a healthy one.
The docs weren't wrong. I'd treated the conservative number like a measurement of my workload.
Then the 500x benchmark failed the same test. I'd measured a sick cluster, then a healthy one.
I reserved 2,887 GiB for metadata. Ceph uses 87. · ProbablyFine
I sized BlueStore metadata for nineteen spinning-disk OSDs at 4% and reserved 2,887 GiB. They use 87.2 GiB, and the benchmark I was about to publish turned out to be a benchmark of my own VM.
probablyfine.dev
September 2, 2026 at 12:01 AM
I sized every Ceph database at 4% and reserved 2,887 GiB of NVMe. Nineteen of them use 87.
The docs weren't wrong. I'd treated the conservative number like a measurement of my workload.
Then the 500x benchmark failed the same test. I'd measured a sick cluster, then a healthy one.
The docs weren't wrong. I'd treated the conservative number like a measurement of my workload.
Then the 500x benchmark failed the same test. I'd measured a sick cluster, then a healthy one.
#553066 syft: 1.50.0 -> 1.51.0
#553065 sqlmap: 1.10.6 -> 1.10.8
#553063 cider-2: 4.0.9.1 -> 4.0.17
#553059 openhack: 0.2.1 -> 0.2.3
#553058 ceph.tests: Fix dashboard test timeout, move it to bluestore test
#553054 multica-cli: 0.4.20 -> 0.4.26
#553048 python3Packages.xknx: 3.18.0 -> 3.19.0
#553065 sqlmap: 1.10.6 -> 1.10.8
#553063 cider-2: 4.0.9.1 -> 4.0.17
#553059 openhack: 0.2.1 -> 0.2.3
#553058 ceph.tests: Fix dashboard test timeout, move it to bluestore test
#553054 multica-cli: 0.4.20 -> 0.4.26
#553048 python3Packages.xknx: 3.18.0 -> 3.19.0
August 16, 2026 at 12:05 AM
#553066 syft: 1.50.0 -> 1.51.0
#553065 sqlmap: 1.10.6 -> 1.10.8
#553063 cider-2: 4.0.9.1 -> 4.0.17
#553059 openhack: 0.2.1 -> 0.2.3
#553058 ceph.tests: Fix dashboard test timeout, move it to bluestore test
#553054 multica-cli: 0.4.20 -> 0.4.26
#553048 python3Packages.xknx: 3.18.0 -> 3.19.0
#553065 sqlmap: 1.10.6 -> 1.10.8
#553063 cider-2: 4.0.9.1 -> 4.0.17
#553059 openhack: 0.2.1 -> 0.2.3
#553058 ceph.tests: Fix dashboard test timeout, move it to bluestore test
#553054 multica-cli: 0.4.20 -> 0.4.26
#553048 python3Packages.xknx: 3.18.0 -> 3.19.0
CVE-2026-43273 - ceph: supply snapshot context in ceph_zero_partial_object()
CVE ID : CVE-2026-43273
Published : May 6, 2026, 12:16 p.m. | 6 minutes ago
Description : In the Linux kernel, the following vulnerability has been resolved:
ceph: supply snapshot context in ce...
CVE ID : CVE-2026-43273
Published : May 6, 2026, 12:16 p.m. | 6 minutes ago
Description : In the Linux kernel, the following vulnerability has been resolved:
ceph: supply snapshot context in ce...
CVE-2026-43273 - ceph: supply snapshot context in ceph_zero_partial_object()
In the Linux kernel, the following vulnerability has been resolved: ceph: supply snapshot context in ceph_zero_partial_object() The ceph_zero_partial_object function was missing proper snapshot context for its OSD write operations, which could lead to data inconsistencies in snapshots. Reproducer: ../src/vstart.sh --new -x --localhost --bluestore ./bin/ceph auth caps client.fs_a mds 'allow rwps …
cvefeed.io
May 6, 2026 at 2:25 PM
CVE-2026-43273 - ceph: supply snapshot context in ceph_zero_partial_object()
CVE ID : CVE-2026-43273
Published : May 6, 2026, 12:16 p.m. | 6 minutes ago
Description : In the Linux kernel, the following vulnerability has been resolved:
ceph: supply snapshot context in ce...
CVE ID : CVE-2026-43273
Published : May 6, 2026, 12:16 p.m. | 6 minutes ago
Description : In the Linux kernel, the following vulnerability has been resolved:
ceph: supply snapshot context in ce...
なんかうまいことした事例はあるっぽいけど
https://forum.proxmox.com/threads/ceph-cluster-rebuild-import-bluestore-osds-from-old-cluster-bad-fsid-osd-dont-start-he-only-stays-in-down-state.151349/
https://forum.proxmox.com/threads/ceph-cluster-rebuild-import-bluestore-osds-from-old-cluster-bad-fsid-osd-dont-start-he-only-stays-in-down-state.151349/
[SOLVED] - ceph cluster rebuild - Import bluestore OSDs from old cluster (bad fsid) - OSD dont start. He only stays in down state
ceph version 18.2.2 (e9fe820e7fffd1b7cde143a9f77653b73fcec748) reef (stable) Hello everyone, please we need your help to import the bluestore OSDs to the new cluster, because after trying to import, we noticed that the OSDs do not start. They are only in a "down" state and are imported as...
forum.proxmox.com
June 25, 2025 at 9:39 PM
Ceph cluster suddenly rebalancing after years of stability? Check three things: 1. PG autoscaler defaults changed during your last version upgrade 2. OSD nearfull ratio crossed as data grew 3. BlueStore migration altered crush rules `ceph config dump | grep autoscal` reveals the culprit most of t...
March 1, 2026 at 11:50 PM
Ceph cluster suddenly rebalancing after years of stability? Check three things: 1. PG autoscaler defaults changed during your last version upgrade 2. OSD nearfull ratio crossed as data grew 3. BlueStore migration altered crush rules `ceph config dump | grep autoscal` reveals the culprit most of t...
Got a bit of curry from `dovecot` storing an email, then saw a error 500 out of Mastodon… went looking:
```
e
2025-07-07T17:16:23.098879+1000 mon.hydrogen [WRN] Health check update: 2 OSD(s) experiencing slow operations in BlueStore (BLUESTORE_SLOW_OP_ALERT)
2025-07-07T17:20:00.000204+1000 mon […]
```
e
2025-07-07T17:16:23.098879+1000 mon.hydrogen [WRN] Health check update: 2 OSD(s) experiencing slow operations in BlueStore (BLUESTORE_SLOW_OP_ALERT)
2025-07-07T17:20:00.000204+1000 mon […]
Original post on mastodon.longlandclan.id.au
mastodon.longlandclan.id.au
July 8, 2025 at 9:56 AM
Got a bit of curry from `dovecot` storing an email, then saw a error 500 out of Mastodon… went looking:
```
e
2025-07-07T17:16:23.098879+1000 mon.hydrogen [WRN] Health check update: 2 OSD(s) experiencing slow operations in BlueStore (BLUESTORE_SLOW_OP_ALERT)
2025-07-07T17:20:00.000204+1000 mon […]
```
e
2025-07-07T17:16:23.098879+1000 mon.hydrogen [WRN] Health check update: 2 OSD(s) experiencing slow operations in BlueStore (BLUESTORE_SLOW_OP_ALERT)
2025-07-07T17:20:00.000204+1000 mon […]