Christoph Lutz @christophlutz.bsky.social · 03/10/2026arxiv.org/abs/2609.22781arxiv.orgio_uring in Oracle Database: A Hybrid Storage I/O Architecture at Production ScaleWe describe the integration of io_uring into Oracle Database's storage layer and the architectural decisions required to deploy it in a production multi-process RDBMS. Our design uses per-process ring... 030
Reposted by Christoph LutzKamil Stawiarski @ora600pl.bsky.social · 23/09/2026Burning tokens for useful toys continues. I enjoy using SQLcl but if you (like me) use oradebug for testing, investigating and mostly fun, you will be forced to use SQL*Plus. So I instrumented Codex to create a plugin for SQLcl to call upidbg, kpusvcrh and kpusvc2hst :) github.com/ora600pl/soda 022
Christoph Lutz @christophlutz.bsky.social · 15/09/2026Interested in how Oracle 26ai protects local auto-login wallets? 👇 t.ly/4re1yt.lyDeobfuscating Local Auto-Login Wallets in Oracle 26aiDeobfuscating Local Auto-Login Wallets in Oracle 26ai 121
Reposted by Christoph LutzMartin Berger (berx) @martinberx.bsky.social · 06/09/2026#POUG2026 is over - post conference blues kicks in. Thank you @luizafrompoland.bsky.social & @ora600pl.bsky.social and all your team for the fabulous event! Big thanks to all the speakers and attendees for content, questions & discussions. You all make this conference unique. See you at #POUG2027 0174
Reposted by Christoph LutzDani Schnider @danischnider.bsky.social · 05/09/2026Some „basics“ about log writer (LGWR) explained by @christophlutz.bsky.social at #POUG2026, including a live demo with foreground session @franckpachot.bsky.social and log writer @chandlerdba.bsky.social 2102
Christoph Lutz @christophlutz.bsky.social · 01/09/2026In Oracle 26ai, auto-login wallets are protected by a randomly generated password encrypted with a random 128-bit AES key. Both are stored in the cwallet.sso header, allowing the original ewallet.p12 and its TDE master key to be recovered. Full details: t.ly/hfZXSt.lyDeobfuscating Auto-Login Wallets in Oracle 26aiDeobfuscating Auto-Login Wallets in Oracle 26ai (cwallet.sso) 000
Christoph Lutz @christophlutz.bsky.social · 12/08/2026www.0xchris.dev/posts/2026-0...0xchris.devDeobfuscating Database Link Passwords in Oracle 26aiDeobfuscating Database Link Passwords in Oracle 26ai 010
Christoph Lutz @christophlutz.bsky.social · 11/08/2026The DBMS_CREDENTIAL.CREATE_CREDENTIAL procedure obfuscates passwords by encrypting them with a randomly generated AES-128 key. The encrypted password and its encryption key are then base64 encoded and stored in the data dictionary. Full details: t.ly/mp5bBt.lyDeobfuscating Oracle Credential PasswordsDeobfuscating Oracle Credentials Passwords (dbms_credential) 000
Christoph Lutz @christophlutz.bsky.social · 31/07/2026Had a closer look at the implementation of the CURSORTRACE event... and encountered a few surprises: t.ly/jQG6Bt.lyUnderstanding the CURSORTRACE EventThe Oracle library cache is extensively instrumented, but the instrumentation is disabled by default. It can be enabled using the CURSORTRACE event, which is particularly useful for investigating libr... 022
Christoph Lutz @christophlutz.bsky.social · 15/07/2026Undocumented SYS_CONTEXT parameters: t.ly/822eRt.lyUndocumented SYS_CONTEXT ParametersUndocumented SYS_CONTEXT Parameters 063
Christoph Lutz @christophlutz.bsky.social · 09/06/2026Ever wondered how exactly Oracle scans child cursors? Going down the rabbit hole 👇 t.ly/5fYsit.lyChild Cursor Scan Mechanics - An InvestigationChild Cursor Scan Mechanics - An Investigation 141
Christoph Lutz @christophlutz.bsky.social · 03/06/2026A short note on how to identify restricted and undocumented oradebug commands. t.ly/QyOFxt.lyRestricted and Undocumented oradebug CommandsRestricted and Undocumented oradebug Commands 030
Christoph Lutz @christophlutz.bsky.social · 01/06/2026The greatest DBA tool of all time is gdb - change my mind. 100
Christoph Lutz @christophlutz.bsky.social · 20/05/2026t.ly/8A1p9t.lyDefault Calculation of TARGET_PDBSAccording to My Oracle Support note PALRT1151 (see also source 1) and various sources on the internet, Oracle uses the following formula to derive the default value of the target_pdbs parameter: targe... 010
Christoph Lutz @christophlutz.bsky.social · 19/05/2026A little script and some background information on how to extract v$ definitions directly from the oracle binary. t.ly/NT6L7t.lyExtracting Fixed View Definitions From The Oracle BinaryExtracing Fixed View Definitions From The Oracle Binary 000
Reposted by Christoph Lutzavelio.ch @avelio.ch · 13/05/2026Avelio is live! 🚀 We are your Swiss partner for Oracle on‑premises database operations. Our mission: Optimised. Automated. Always Available. From proactive monitoring to 24/7 SLA‑backed support. Your business deserves reliability. Find out more: avelio.ch #Oracle #DatabaseManagement 094
Christoph Lutz @christophlutz.bsky.social · 07/05/2026A closer look at how Oracle 26ai determines and caps the maximum granule size t.ly/MEKOUt.lySGA Changes in Oracle 26ai - Maximum Granule SizeSGA Changes in Oracle 26ai - Maximum Granule Size 030
Christoph Lutz @christophlutz.bsky.social · 05/05/2026t.ly/FFWYdt.lySGA Changes in Oracle 26ai - Subpools and DurationsSGA changes in Oracle 26ai - subpools and durations 085
Christoph Lutz @christophlutz.bsky.social · 20/04/20261/2 Interesting and undocumented changes in SGA management in Oracle 26ai: - Up to 31 subpools in the shared pool (max seven in 19c defined by _kghdsidx_count) - Subpools are split into thee durations (there were only two in 19c) 162
Christoph Lutz @christophlutz.bsky.social · 14/04/2026t.ly/iOYIxt.lyBypassing oradebug RestrictionsBypassing oradebug restrictions 020
Christoph Lutz @christophlutz.bsky.social · 18/03/2026Pipelined or parallel? In Adaptive Scalable mode LGWR can assign redo writes to worker processes in different ways. More details in the post below 👇 t.ly/HJi5dt.lyAdaptive Scalable LGWR Pipelined vs Parallel Redo WritesAn earlier post explained the decision-making heuristics when LGWR transitions from single to scalable mode. Here, we examine how it handles redo writes in scalable mode with multiple LG worker proces... 011
Christoph Lutz @christophlutz.bsky.social · 04/03/2026blog.tohojo.dk/2026/02/the-...blog.tohojo.dkThe inner workings of TCP zero-copy 000
Christoph Lutz @christophlutz.bsky.social · 03/03/2026When does lgwr transition from single to scalable mode? t.ly/Ohe7at.lyAdaptive Scalable LGWR Mode Switch Threshold Single->ScalableOracle fundamentally reworked and redesigned the LGWR architecture in 12c. In earlier versions, LGWR ran as a single process, whereas Oracle 12c and later can dynamically switch between a single LGWR ... 023
Christoph Lutz @christophlutz.bsky.social · 13/02/2026Everybody should get rid of tnsnames.ora and use Easy Connect instead. Change my mind. 030
Christoph Lutz @christophlutz.bsky.social · 10/02/2026Just discovered this new blog series by @ludovico.bsky.social explaining Data Guard enhancements in 26ai - must read👇 www.ludovicocaldara.net/dba/dg-26ai-...ludovicocaldara.netMini-blog series: Oracle Data Guard 26ai new featuresOracle AI Database 26ai is here for Linux x86_64—packed with new features for Data Guard! Stay tuned for quick blog posts on key changes. 051
Christoph Lutz @christophlutz.bsky.social · 12/01/2026I felt like taking a closer look at this new datapatch feature and ended up making some interesting discoveries 🙂 t.ly/Tufobt.lyDatapatch And The "force_terminate_blocking_sessions" Option — A Look Behind The ScenesIf you’ve ever executed Datapatch on a busy system, chances are you’ve experienced intermittent delays or hangs, and in some situations maybe even errors such as ORA-4021 (“timeout occurred ... 171
Christoph Lutz @christophlutz.bsky.social · 14/12/2025rovarma.com/articles/fro...rovarma.comFrom profiling to kernel patch: the journey to an eBPF performance fix | Ritesh Oedayrajsingh VarmaA story about how an innocent profiling session led to a change to the Linux kernel that makes eBPF map-in-map updates much faster. 000
Christoph Lutz @christophlutz.bsky.social · 10/12/2025Woot! By a happy twist of fate, a ticket for the 39th Chaos Communication Congress came may way ... 😀 See you in Hamburg! @ccc.de Thanks @krischan.bsky.social! 120
Christoph Lutz @christophlutz.bsky.social · 28/11/2025justoffbyone.com/posts/math-o...justoffbyone.comThe Math of Why You Can't Focus at WorkInterruptions, recovery time, and task size: three numbers that determine if you'll get real work done. Interactive visualizations show the math behind bad days. 100
Christoph Lutz @christophlutz.bsky.social · 23/11/2025#POUG2026 dress code 😜 www.galaxus.ch/en/page/woul... @pougorg.bsky.social 180
Christoph Lutz @christophlutz.bsky.social · 05/11/20251/11 Everything changes... turns out the age-old rule that lgwr writes out the log buffer when it's 1/3 full no longer applies in recent Oracle versions. Observations below from 19.26 (with RAC on Exadata). 👇 120
Christoph Lutz @christophlutz.bsky.social · 31/10/2025Week end fun: snooping inter-process messaging (ksbasend) in Oracle with bpftrace. 🤓 t.ly/_yDy7 020
Christoph Lutz @christophlutz.bsky.social · 20/10/2025thehackernews.com/2025/10/link...thehackernews.comLinkPro Linux Rootkit Uses eBPF to Hide and Activates via Magic TCP PacketsSynacktiv uncovered LinkPro, a Golang rootkit using eBPF hide/knock modules activated by TCP window 54321. 000
Christoph Lutz @christophlutz.bsky.social · 11/10/2025www.deep-kondah.com/deep-dive-in...deep-kondah.comUnraveling eBPF Ring BuffersThe goal of this post is to provide an in-depth discussion of BPF ring buffers, covering their internals, including memory allocation, user-space mapping, locking mechanisms, and efficient data sharin... 010
Christoph Lutz @christophlutz.bsky.social · 05/10/2025When you plan to geek out over some oracle internals, but end up ftrace’ing bpf the entire week end to chase a funny bug that only occurs on exadata with capacity on demand ... 010
Reposted by Christoph LutzJonathan Lewis @jloracle.bsky.social · 28/09/2025There's a problem on the oracle-l listserver at present about an insert taking far too much time (and CPU). It's a known issue and there are 47 statistics in v$sysstat (19.11) with names like 'ASSM%' to help diagnose it. How many do you think are described in the database reference manual? None. 041
Christoph Lutz @christophlutz.bsky.social · 21/09/2025Yet another adaptive lgwr optimization: on Exadata X10+, pipelined log writes may defer redo writes until a suitably sized write batch has accumulated in the log buffer. The deferral can involve spinning in a tight loop up to 25 times (maximum hard-coded in kcrfw_defer_write). 120
Reposted by Christoph LutzSwiss Oracle User Group @soug.ch · 10/09/2025In den nächsten Tagen veröffentlichen wir nicht nur die Agenda, sondern am Tag nach dem SOUG Day planen wir noch ein Special für Euch! Schaut rein und meldet Euch an unter soug.ch. 033
Christoph Lutz @christophlutz.bsky.social · 04/09/2025So glad that all new features are documented so well... NOT 😜 Manually enabling and disabling adaptive lgwr evaluation trace for pipelined / overlapped redo writes: 000
Christoph Lutz @christophlutz.bsky.social · 03/09/2025POUG journey started… not even at the airport and lufthansa’s delay notification leaves no hope of making the connecting flight in MUC 🙈 030
Christoph Lutz @christophlutz.bsky.social · 31/08/2025So 23ai replaces ksesecl0(func, loc, err) with kseseclv(err, func, loc, ...) ... Why is it always just a few days before POUG that this kind of low-level discoveries surface? 🙄 230
Reposted by Christoph LutzTanel Poder @tanelpoder.com · 13/08/2025A new tool in 0x.tools family: xtop - Top for Wall-Clock Time. It uses eBPF/xcapture v3 and gives you "x-ray vision" into Linux system activity. It will be available on next Tuesday 19 Aug at 1pm EDT when I also run a live demo webinar! tanelpoder.com/posts/xtop-t... 1199
Christoph Lutz @christophlutz.bsky.social · 12/08/2025"Slide n of 142".. this is getting out of control... 🙈 010
Christoph Lutz @christophlutz.bsky.social · 06/08/2025Today's discovery (19.26): Oracle derives different log parallelism defaults, depending on platform. Max number of public redo strands is: Exadata: CPU_COUNT <= 256: 16 CPU_COUNT> 256: CPU_COUNT/16 Non-Exadata: CPU_COUNT <= 32: 2 CPU_COUNT > 32: CPU_COUNT/16 Max limit: 256 000
Christoph Lutz @christophlutz.bsky.social · 05/08/2025Oh my, looks like the "ms" in "kso_sched_delay_avg_ms" actually means "microseconds" ... 🤷♂️ 130
Christoph Lutz @christophlutz.bsky.social · 04/08/20251/6 The "redo synch time overhead" in Oracle is the difference between a FG's log file sync (LFS) wait end time and LGWR's redo write completion time. LGWR and LG workers track the redo write completion times in the "write info array" in the SGA. 100
Christoph Lutz @christophlutz.bsky.social · 03/08/2025lgwr: You have the control gdb: I have the control 000
