Cache replacement policies Puns

Hi everyone,

So recently I changed my Cache drive and went through the normal procedures (move files onto Array, replace drive, move files onto Cache).

After doing that however, all of my dockers work except for Plex. Any ideas as to what could have gone wrong?

I tried rebuilding from Docker image as well, and the same issue.

Plex was working fine right before changing the Cache drive as well.

I have never had any issue with Qrates, I have received 3 albums so far in great condition. Unfortunately, I got the newest Fearless Flyers record and the top of the album sleeve had begun to split.

Qrates promises a refund or replacement and I just wanted to hear about anyones experiences with this!

I implemented an approximation to LRU in javascript and tested reads and writes in isolation and they do approach expected percentages of hits or misses. But when I use both reads and misses equally in a benchmark with random access, read cache hit stays at expected value but the write cache hit approaches to the read hit squared.

Example:

0.5 read hit, 0.25 write hit

0.8 read hit, 0.64 write hit

0.99 read hit, 0.98 write hit

I really have no idea where I did something wrong but cache just works for any access pattern (z-order, sequential, random, etc) good enough but write hits always stay a bit lower.

So, I'm just asking if some types of caches are naturally having unbalanced reads writes efficiencies, performances for specific acceess patterns what could be a good way of benchmarking and is there an article that experiments such differencies?

This is kind of a rant. I ordered 4 candle holders during the B2G2 sale. I got my order yesterday and they only shipped me 2 of the holders. I called and was told the 2 that didn’t ship are out of stock. I told the girl they are still on the website and she said they are below what they can use to fill replacement orders. But this wasn’t technically a replacement order because they never shipped 2 of them, so their stock should be up 2. They did end up shipping me 2 new holders because I will say B&BW Customer Service has always been top notch, but I just don’t understand why they can’t ship me what I originally ordered. I guess I’m just frustrated because something has been wrong with the last 4 orders I have placed.

I have a significant amount of dust under my photo lens (seems to be an ongoing issue with the Pixel 4a 5G) that all my photos seem milky.

I contacted Google support (Germany) about a replacement but they tell me that they just changed their policy and I will have to send in my phone for an assessment, all the while having no replacement phone.

Turnaround is estimated at 10 days minimum. Even a day without a phone is borderline impossible, let alone 10.

In the past (I've had google phones since Nexus times), you received a replacement/loaner Phone, which was obviously the more consumer friendly option.

Has this policy change officially been announced somewhere? What's up with being so consumer unfriendly?

PLS READ THE EDIT FIRST :)

I stoped the Docker Service, set all shares to cache: yes. It worked first but now I still have 55GBs on the cache. Some Appdata and The dockerimage. The mover will not move anymore.. What now?

I have a 3 days old appdata backup from the plugin. What would be the fastest way to delete the moved files from the array and replace the cache drive and the nrestore from the plugin?

Thanks in advance

EDIT: I changed the cache drive now and read the files from the old cache through unassigned devices. The Appdatafolder were all emty, the mover forgot just to delete all the folders... there is only the docker image. can I simply copy past this before reenable the dockerservice first?

Haul pending for over 90 days. Supposedly shipped but never received. BMLin won’t refund or replace per its own policy. Just beware of lost shipments and bad service from BMLin.

ETA2: Wait, I take it back. Host says it's definitely running Apache and "nginx is just a proxy that runs alongside Apache and manages some of the workload."

ETA1: Nevermind. It was a PEBKAU error. I feel like a dumbass - turns out we're using nginx, not Apache. Doh! In my defense, all the settings in the host panel said Apache, and there was no mention of nginx anywhere. It was only when I checked the Network tab in Dev Tools that I saw it was actually nginx.

I have various DEFLATE, ExpiresByType and Header set Cache-control in my htaccess file, but I'm still getting suggestions to "Serve static assets with an efficient cache policy" from PageSpeed and GTMetrix, etc. Is 1 year not enough for images, etc? What have I done wrong here?

Here's my htaccess:

# BEGIN DEFLATE COMPRESSION
<IfModule mod_deflate.c>
  # Compress HTML, CSS, JavaScript, Text, XML and fonts
  AddOutputFilterByType DEFLATE application/javascript
  AddOutputFilterByType DEFLATE application/rss+xml
  AddOutputFilterByType DEFLATE application/vnd.ms-fontobject
  AddOutputFilterByType DEFLATE application/x-font
  AddOutputFilterByType DEFLATE application/x-font-opentype
  AddOutputFilterByType DEFLATE application/x-font-otf
  AddOutputFilterByType DEFLATE application/x-font-truetype
  AddOutputFilterByType DEFLATE application/x-font-ttf
  AddOutputFilterByType DEFLATE application/x-javascript
  AddOutputFilterByType DEFLATE application/xhtml+xml
  AddOutputFilterByType DEFLATE application/xml
  AddOutputFilterByType DEFLATE font/opentype
  AddOutputFilterByType DEFLATE font/otf
  AddOutputFilterByType DEFLATE font/ttf
  AddOutputFilterByType DEFLATE image/svg+xml
  AddOutputFilterByType DEFLATE image/x-icon
  AddOutputFilterByType DEFLATE text/css
  AddOutputFilterByType DEFLATE text/html
  AddOutputFilterByType DEFLATE text/javascript
  AddOutputFilterByType DEFLATE text/plain
  AddOutputFilterByType DEFLATE text/xml
  # Remove browser bugs (only needed for really old browsers)
  BrowserMatch ^Mozilla/4 gzip-only-text/html
  BrowserMatch ^Mozilla/4\.0[678] no-gzip
  BrowserMatch \bMSIE !no-gzip !gzip-only-text/html
  Header append Vary User-Agent
</IfModule>
# END DEFLATE COMPRESSION


# BEGIN SERVE STATIC ASSETS WITH AN EFFICIENT CACHE POLICY
<IfModule mod_expires.c>

# Enable expirations
  ExpiresActive On

Hi po may question lang po ako kasi bagong ps4 owner lang po ako. Medyo late na I know so very recently nagkaroon po ng sira yung console ko just after 4 days of careful use nag shushutdown po sya on it's own kahit meron namang power sa led light po nya. Obviously since pasok po sya sa 7 days warranty period ng datablitz nag report po ako immediately at gagawan daw po nila ng assessment and if valid irereplace nila. May 2 questions po ako na gusto ko po sana ma clarify ng mga veteran na or may similar experience.

1. Possible po ba na refurbished yung console na nakuha ko? Napansin ko kasi na last unit na sya sa physical store. At napaka suspicious ng kaunti na may issue pa yung nakuha ko.

2. If ever na na experience nyo na po magpa replace ng console, gaano katagal nyo po nakuha or na deliver sa inyo yung replacement? Medyo anxious na po kasi ako makapaglaro ulit.

Thank you po in advance

Over the last 6 months or so, I see a growing trend of issues where the user didn't have a latest group policy. The fix is simple, just delete the contents c:/windows/system32/GroupPolicy on the computer, gpupdate and and it re-downloads everything and fixes the issue.

Anyone else have these kind of issues? We're running DC's 2019, and workstations at least Win 10 Enterprise 1909.

Thanks.

I understand you can stop the Docker/VMs and invoke the mover to move files onto the array while you swap out the physical drives. Is there a way to just move the files from the current SSD to an unassigned drive SSD (the new one)? What would you use to move the files? I assume it would be faster since it's SSD to SSD.

They can get around their customer service phone maze and get your money back

The Firestore web SDK has support for offline mode with snapshots which is amazing when the cache is small but if it grows large it gets VERY slow.

Our app takes about 3.5s to load now because of the performance issues.

Every snapshot requires an O(N) scan of all the data. There aren't any indexes used.

I have a prototype implemented, and it seems to be about 7x faster than the current cache.

I tried to reach out to the Firestore team about this a few times and the general consensus is that they aren't going to fix this bug anytime soon.

I realize they have a lot on their plate bu this is really hurting my company so I need it fixed.

The cache I'm working on is designed around a limited view of the Firestore data. Firestore probably, itself, can't be easily cached without a ton of work but for a lot of use cases, if you use a limited API (still about 80-90% of what Firestore offers) then it should work fine.

The general idea is that there are document and query snapshots.

A query snapshot can return multiple documents and a document snapshot is obviously just one document.

The idea is to support storing the docs individually in the cache then having secondary indexes representing the query snapshots.

So a secondary index would just be the list of document IDs in that snapshot, in the order they were returned.

I have an initial proof of concept here:

https://github.com/burtonator/polar-app-public/tree/master/polar-snapshot-cache

and it's already much faster.

Right now I'm storing a full materialized view of the query snapshots but I want to investigate if it's faster to just store the index and the individual documents directly.

I'm also working on some 'modes' for the cache such that you can measure cache performance, hits, etc to verify that it's working properly when you turn it on. Also, it would be nice to have a 'wrapped' mode that could compute cache stats so we could compare it against the stock Firestore cache.

The idea is that we wrap the normal Firestore object with a decorator which has a more limited API surface area.

I'm also going to be working on some optimizations to potentially avoid cache updates for every snapshot such that if the objects haven't changed, even though they're in a snapshot result, we don't update the store. This is only really an issue for the FIRST snapshot since after that Firestore sends you doc changes and you can just see which ones are updated.

... initial load time is super important t

I have created a script which replaces some of the ~/.cache folders with a RAMDisk.

The intended behavior:

• User logs in (graphically or via terminal), the RAMDisk is created, before the desktop starts using the cache folders.
• User logs in more simultaneous sessions: Don't run the script since it's already mounted.
• User logs out from their last remaining session (only do this when all of their sessions are closed): Unmount the RAMDisk and restore the original folders.

I am curious what the correct systemd target is, for achieving this?

An elderly neighbor of mine (we're in California) accidentally dropped her Safelink Wireless phone on the brick floor and severely cracked the screen. I was helping her research via Google, and found several Web sites mention Safelink having a "one-time replacement for free" policy (here's one example). But when she contacted customer service, she was told that since the phone is still usable, they would not approve a replacement, and she would have to handle any repairs on her own.

Does anyone know more details about the supposed "one-time replacement" policy for Safelink? If we can find official evidence of this policy, we will try calling again. So far I've had no luck finding any helpful info on the Safelink Web site, and we had a really hard time understanding the customer service agent over the staticky phone line. 😕

We got a quote of $125 from a repair shop to replace the screen, so if the free replacement route doesn't work out, I think it would make more sense to buy a different phone that would work on Safelink's network rather than repairing the broken screen on the current phone (Alcatel A503DL).

I'd appreciate any ideas or suggestions.

is dell replaced laptop new or refurb ?]

How long does it take get a new replaced unit ?

Does the replaced unit have same or lower hardware ?

if anybody get their system replaced, how was your experience, did you get same or lower hardware ?

Please help.

Hello, I had a failed cache disk on s2d node and have now finally replaced it. The cluster after some time optimized the storage, and I was expecting the cache disk to show with "usage" of "journal" but it still states as "Automatically selected". Now I don't know if the disk is used as a journal disk, or if its simply in the data parity storage..... The replaced disk is an ssd, while other disks are spinning disks. I have two ssd and four spinning disks in each node. How can I identify what the new disk is doing, and do I need to change its role? Thank you

I'm searching in the privileged and unprivileged specs about cache coherence and which cache policy RISC-V implements, but I have one question:

- Does RISC-V (as an ISA) defines any cache coherence policy at all?

I read about PMA's and how they define memory regions as cacheable, but I'm not sure if that defines a coherence scheme or if it is a base structure for a hardware/software cache coherence implementation (i.e, snooping/directory).

ETA2: Wait, I take it back. Host says it's definitely running Apache and "nginx is just a proxy that runs alongside Apache and manages some of the workload."

ETA1: Nevermind. I feel like a dumbass - turns out we're using nginx, not Apache. Doh! In my defense, all the settings in the host panel said Apache, and there was no mention of nginx anywhere. It was only when I checked the Network tab in Dev Tools that I saw it was actually nginx.

I have various DEFLATE, ExpiresByType and Header set Cache-control in my htaccess file, but I'm still getting suggestions to "Serve static assets with an efficient cache policy" from PageSpeed and GTMetrix, etc. Is 1 year not enough for images, etc? What have I done wrong here?

Here's my htaccess:

# BEGIN DEFLATE COMPRESSION
<IfModule mod_deflate.c>
  # Compress HTML, CSS, JavaScript, Text, XML and fonts
  AddOutputFilterByType DEFLATE application/javascript
  AddOutputFilterByType DEFLATE application/rss+xml
  AddOutputFilterByType DEFLATE application/vnd.ms-fontobject
  AddOutputFilterByType DEFLATE application/x-font
  AddOutputFilterByType DEFLATE application/x-font-opentype
  AddOutputFilterByType DEFLATE application/x-font-otf
  AddOutputFilterByType DEFLATE application/x-font-truetype
  AddOutputFilterByType DEFLATE application/x-font-ttf
  AddOutputFilterByType DEFLATE application/x-javascript
  AddOutputFilterByType DEFLATE application/xhtml+xml
  AddOutputFilterByType DEFLATE application/xml
  AddOutputFilterByType DEFLATE font/opentype
  AddOutputFilterByType DEFLATE font/otf
  AddOutputFilterByType DEFLATE font/ttf
  AddOutputFilterByType DEFLATE image/svg+xml
  AddOutputFilterByType DEFLATE image/x-icon
  AddOutputFilterByType DEFLATE text/css
  AddOutputFilterByType DEFLATE text/html
  AddOutputFilterByType DEFLATE text/javascript
  AddOutputFilterByType DEFLATE text/plain
  AddOutputFilterByType DEFLATE text/xml
  # Remove browser bugs (only needed for really old browsers)
  BrowserMatch ^Mozilla/4 gzip-only-text/html
  BrowserMatch ^Mozilla/4\.0[678] no-gzip
  BrowserMatch \bMSIE !no-gzip !gzip-only-text/html
  Header append Vary User-Agent
</IfModule>
# END DEFLATE COMPRESSION


# BEGIN SERVE STATIC ASSETS WITH AN EFFICIENT CACHE POLICY
<IfModule mod_expires.c>

# Enable expirations
  ExpiresActive On

# Default expiratio