dormando

9 ms·

31.
by dormando 7y ago
Thanks!
32.
by dormando 7y ago
Mind giving some examples in more detail about how you make use of lists/scores?
33.
by dormando 7y ago
They make those too. HyperRAM is pretty popular for small FPGA projects: https://www.mouser.com/new/issi/issi-hyperram/ - there are tradeoffs in potential performance/interface/usability/etc bu
34.
by dormando 7y ago
You can try, but it's not going to be very consistent. You're at the mercy of whenever the OS decides to start flushing pages.. head of line blocking, mmap_sem locks, file descriptor locks, etc. My old job had an mmap-on-disk stor
35.
by dormando 7y ago
In this case, if you're using tmpfs like we recommend the performance is identical. It's the same as if we used normal shared memory (or so far as I can tell via benchmarks). Even with an optane pmem mount the perf. is super close
36.
by dormando 7y ago
Yup, but you can copy that file (+ the .meta file) to disk, reboot, copy back, etc. It'll take a long time for a large cache unless your disk is fast. If you have some kind of pmem device and a dax mount, those can survive reboots on t
37.
by dormando 7y ago
That's exactly what it does, though since it's resuming a monotonic clock the actual code was a bit more complicated... To find out how long it was down, it notes system time into the state file on shutdown. On start it checks the
38.
by dormando 7y ago
Before memcached had a monotonic clock people would end up with immortal objects (underflowed TTL's) because ntp would start after memcached and make a huge adjustment due to the hardware clock being really off. With the restart code,
39.
by dormando 7y ago
You can do it at really low load levels, but you'll lose performance consistency. mmap'ing files over a real filesystem of any kind is super complicated. The access pattern isn't optimized at all for flash or HDD or etc... ho
40.
by dormando 7y ago
The kindle version apparently used to be a shoddy PDF scan. Any idea if amazon's been updated to the 2018 reissue and if that's any better? All of the reviews claim it's crud.
41.
by dormando 7y ago
Years ago I really wanted the sauce to AOL's NWN. Gold Box RPG in MMO form! I actually played it online back when AOL was $5/hr (after the first five free hours!). At some point I talked to someone who worked at the co who ate the
42.
by dormando 7y ago
I keep getting raged at over e-mail/github because end-packagers and some users don't comprehend why it's broken. I completely regret adding seccomp support in the first place, since it breaks randomly, packagers enable it by
43.
by dormando 7y ago
I had a ton of fun with XBAND until people discovered *69. parents were not amused when rando's kept calling to "see what the other guy sounded like". Drooled constantly over the XBAND keyboard that they probably only made te
44.
by dormando 7y ago
People learn and re-learn this at almost every company, which I find fascinating. The redis nx bits aren't a bad way of dealing with it. Since they mention memcached: I've been working on a protocol extension to bake this exact th
45.
by dormando 7y ago
Looks like there's a 2nd ed (mips based?) and an "arm edition" (more recent?) of the same title by the same authors. Online reviews are sparse, wonder which is better now?
46.
by dormando 7y ago
TinyFPGA BX is pretty good (all open source tooling), but generally everyone has linux tools. edit: unless you meant which one would boot linux best :P Answer is "the expensive ones"
47.
by dormando 8y ago
I did most of one of these on an FPGA in verilog for fun: https://github.com/dormando/verilog-raycaster - took a lot longer than a weekend :)
48.
by dormando 8y ago
The article was significantly cut down for brevity, so any detail of how to find the right balance is beyond the scope :) The original question was that it can be really hard to determine how much adding RAM will offload from the disk. IE:
49.
by dormando 8y ago
Ah my point of reference was more "how do you go from memcached RAM to RAM + disk?" not "how do you go from no memcached to memcached". If you don't have any cache layer at all, that's a standard approach.
50.
by dormando 8y ago
1) Super workload dependent unfortunately. Probably "easiest" way is to set up test systems with enough RAM and disk space. Then either slowly reduce the RAM or start from a low point and slowly increase until you hit your target;
51.
by dormando 8y ago
close enough. heavily batched on a 48 core I can pull 50 million keys/sec over localhost. if you remove syscalls and use it as a library it should double at least. writes are another story, but they're slower because nobody asks f
52.
by dormando 8y ago
heh.. think I had a rack near prgmr at svtix. I used to play a game with seeing how little information I could give them before they unlock my rack (and usually different people working there) 1) ID, rack hall, rack number, company name. 2)
53.
by dormando 8y ago
most (normal?) datacenters have A/B/C power feeds. Also east/west network runs. even if you don't have multi-line PSU's you can at least stagger the racks. part of being cut-rate is that they don't have these t
54.
by dormando 8y ago
no reason, it's doable with a bit more cleanup. The I/O bits are encapsulated entirely in extstore.c with a relatively clean interface. it's not a layered setup, it's actually bolted all into the same hash table, which m
55.
by dormando 8y ago
Much of the items that will end up on an SSD are already long lived. Every gigabyte of RAM runway before flushing items reduces the write load, since recent items are at highest risk of being overwritten/deleted.. also stuff with short
56.
by dormando 8y ago
They have quite a ways to go before I'd be happier treating them as RAM for something like memcached... but it's a good start. The IOPS is very low and the latency is still quite a ways off from RAM. For a lot of other workloads i
57.
by dormando 8y ago
Please don't try to dial this back to win. I showed you how things work, go ahead and fiddle with them as you want. 1) sure, 100b, but that will just make it easier for the CPU version to hit the packet rate limit. I dialed it down to
58.
by dormando 8y ago
mis-threaded my response below (didn't have a reply button?), so see that too. Just signed up for a personal AWS account and manually started an r4.4xlarge for target and c5.4xlarge for source (same CPU's and networking capability
59.
by dormando 8y ago
Try 12 workers to start, there are some bg threads. I've plenty of experience with both hardware and virtual machines, I just don't use AWS myself much. I can get something like 800k read ops/sec from a 4core raspberry pi2, a
60.
by dormando 8y ago
They stopped updating it a few years ago; it's probably also not as well tuned as you think. I'd need to see the output of "stats settings" from a running instance to know for sure. I also have no idea if it's a hac

More ›