Saturday, May 07
Down Is The New Up Edition
- Xbox is was down. (Bleeping Computer)
It looks like it came back online just minutes ago, after about 11 hours during which you couldn't play games you owned on the console you owned because the truth is, you don't.
- QNAP. (Bleeping Computer)
Again. Though this one seems specific to its security video systems and not to its generic NAS devices.
- I have 100,000 unread emails. What do? (ZDNet)
The answer is "declare email bankruptcy" and archive all of them.
- Conjunction disfunction: These five magic words crash Google Docs. (Bleeping Computer)
The five words were And. Type that five times in a row - And. And. And. And. And. And. - with grammar suggestions turned on, and and and and and splat.
Sadly it seems to have been fixed now.
- Apple's M1 and A14 CPUs have a speculative prefetch bug similar to Spectre and Meltdown. (Tom's Hardware)
- You wouldn't buy a book, would you? (Ars Technica)
Well, now you can't. The Kindle app for Android currently allows you to purchase Kindle books right in the app with a single click - great if you're reading through a series, because it shows up when you get to the last page.
That's going away because Google wants a cut. Instead you'll have to open Amazon in your browser.
- Amazon is planning to spend $12 billion on five new datacenters in Oregon. (The Register)
This is expected to create 600 local jobs - which is not a lot. Most of the complicated stuff is done remotely; the local staff are responsible for putting things in racks and then taking them out again a few years later.
- Intel will be bringing back high-end desktop systems with Sapphire Rapids. (WCCFTech)
Based on the current Alder Lake cores, but with more of them - up to 24 on the mainstream range, and up to 56 cores per socket - and two sockets - on the expert range. Clock speeds up to 5GHz, DDR5 RAM, PCIe 5 I/O, and TDPs up to 500W.
Time to get that 2000W power supply.
Expected in Q3 this year, which is not that far off.
I recently bought two Western Digital 4TB "blue" hard drives. I formatted them (gpt partition table, ntfs partition) and copied over data from two drives that I suspect might be going off (louder and more frequent drive noises.)
Weird thing: Before I moved the data over, the WD drives acted fine. After I moved about 1 TB of data over, both WD drives are making constant head-motion noise, even when I can't imagine they would be doing anything. (When powered on and idling at the BIOS screen, for example - no operating system could possibly be giving them commands!)
What the heck are they doing? It's both of them, so if it's a defect, it's a defect with both of them. Is the drive firmware trying to move blocks around or something? Why would it have been copied over in a fragmented manner, such that it needs defragging?
Until I figure it out, can't really trust the replacement drives! They'll burn themselves out if they don't stop doing this.
Posted by: madrocketsci at Saturday, May 07 2022 11:44 PM (hRoyQ)
Check the SMART stats under "193 Load_Cycle_Count" (I believe). Most drives are rated at around 500k cycles, and when I still used WD garbage, those would go up by 30-50k/day as the drive automatically parked the heads every couple of seconds, and the OS would wake it up right after that with an IO request.
Also, check if it's an SMR drive: It may be simply rewriting everything because the HDD designers are certified retards of the first rank.
Posted by: normal at Sunday, May 08 2022 03:30 AM (obo9H)
Posted by: Mauser at Sunday, May 08 2022 06:32 PM (gVjvf)
Posted by: Rick C at Monday, May 09 2022 01:47 AM (6oqhX)
58 queries taking 0.248 seconds, 320 records returned.
Powered by Minx 1.1.6c-pink.