Hi all, hard week for me. Been sick almost all week and family too, so not much time to play around with stuff.
But still, I managed to get something done.
Fixed my hardware software issues
On the system I am running things from I had this old monitoring software which I won't say the name, that was causing problems with my new GPU, causing 8 to 24 hours crashes. I have now moved into something that is not yet to my liking but looks like doing the job within my minimal requirements.
Next will be, maybe in 2 or 4 weeks (depending on time) to get more 32GB of RAM and expand my setup.
Meanwhile...
I have been playing with these boys, under VMs! You would not believe the amazing stuff they can do in virtual environments (aka, being used by multiple VMs for different things, such as cache LVs under LVM).
I wanna try some BeeGFS things too but still need to experiment more to get a feel for the work/benefit requirement. Then I can probably share some guide or something if there is interest in it.
These setups (using parallel non-block based IO layers like BeeGFS) are especially useful for small setups where you can control a bit of the hardware or the number of volumes you can get (if you are under a cloud environment). They squeeze all the juice from parallelism and provide great flexibility on things you usually needed otherwise to stop/reboot/restart.
Plus, look at that BeeGFS logo, a totally appropriate selling point for a HIVE post! I have to think about it.
Any updates on the HE-AWM script?
Not really... not something you need to check if you are already using it. 🤣 I am so eager and naive at this point, thinking someone is using it LOL.
Either way, if you really are, thanks for trying it out. And I would really appreciate some of that thing everyone calls it, feedback? Heheheh
Don't be afraid of being a critic or rude, I have already some years of trashcan under my chest, and I have a pretty great idea of how terrifying (delivering something done in bash) can this be for others... but I just can't help it. 🙃
I just made some minor commentary updates as I thought fit and prepared something for when I release the next update. I tagged this one v1.1a (keeping letters for minor changes or documentation updates that don't cause behavior changes).
Hive-Engine Snapshots
Finally and because I did a few mistakes during the last days (can happen to anyone, but I think it was my sick head doing the job), things did not go so fast. And if it was not for @rishi556 snapshots shared in his post, I would be extending my synchronizations a few days more.
So, and with the idea of extending his efforts, I have just shared a directory from @forkyishere google drive, and I will be updating that one with some regular (or by request if I am around) snapshots dumps from the MongoDB. Currently, the dump uses just a tinny bit more than 4GB, so up to the limit of the drive (15GB), I can continue this very easily. Afterward, we will see.
Here is the Google Drive link!
Anyone with the link should be able to read the folder contents and download the snapshot/dump archives.
I use --iso-8601 dates and the block number after. The date is from when the dump was taken and might not always correspond to the date the chain was live on that block number.
Any problems you may find with any of the snapshots, please report back (tag @forykw) so I can update the description of a document I also set there ("Snapshots - Issues Report"), in case someone else has used the same snapshot and wishes to get more info.
Next?
I will be going up and down on the node because I have a lot to catch up on at real work. So, depending on that I might or not have the node running all time.
It's amazing how small the M.2 drives really are. When I opened mine up for my PC I was really surprised by it. Glad you found my snapshots helpful, have them up for this exact reason, syncing takes forever. I really want to get myself either a in home server or a server colocated close to home to have something to tinker with, but being a broke college student really doesn't help(nor does it help I'm in 2 different places half the year).
Yeah, I know the feeling. I spent a great part of my university time on unuseful trips. If it was today, I would have moved next to the study place and lived there.
These Samsung NVMe's are not super crazy in random IO's (especially not super on reads in low queue depths, which is what I am looking for), but on throughput they are massive! The PCIe 4.0 new version does around 6.4 GB/s on windows! It's crazy... a quick comparison on windows (with the SAMSUNG Magician app):
These random IO/s are not very representative because they are with 2 queue depth if I am not mistaken... which is stupid to report in a tool like this for an NVMe... But I understand why they do report low queue depths instead.
Now, the moment you increase queue depth to 10's for example... this thing becomes very interesting on IO. Will try to do a more comprehensive benchmark from within a VM once I can, which is what I am exploring. On windows, this is overkill for what most people do.
Just the way we like it. It's amazing how cheap it's getting over time. You can now get over 1 TB SSD(NVME too) for under $100(though you won't get those speeds). I'm hoping for days where the smallest SSDs start at 1 TB because they get that cheap.
Yeah... the EVO was 2x the price of the PRO, 2 years ago. 😭
At work, I deal with enterprise 2 year old SSDs and NVMe's... 8TB and 14TB for the SSD ones... 400GB for the NVMe's (but crazy fas on IOs, around 100k's per level of queue depth almost) which interface with 12Gb/s SAS and they can't reach these sequential throughputs, not even half most likely. So definitely now it's going to get cheaper for general market consumption.
I am eager to start seeing that 400 TB and above SSDs that got announced (prototype) in 2019. When those reach the enterprise market, this is going to be interesting. Probably on my next HPC cluster.
400 TB... Holy hell. How much do they cost? And more importantly, how long till they become suitable for consumer purchases? My 256 GB ssd on my laptop will be puny compared to that.
https://www.techradar.com/nz/news/first-petabyte-usable-capacity-ssd-could-come-in-2023
Time to sleep till 2023. @deathwing buy me one of those when they launch please.
Congratulations @atexoras! You have completed the following achievement on the Hive blockchain and have been rewarded with new badge(s) :
Your next target is to reach 300 upvotes.
You can view your badges on your board and compare yourself to others in the Ranking
If you no longer want to receive notifications, reply to this comment with the word
STOPJust realized now that the google drive link was missing. Post updated with the link now.
Aww, when I got the tag I thought I had a brand new interesting post to read.
You will soon. Sorry, fixing mistakes on the post. Preparing the new post...
Because somehow I would love to be in your position...
!WINE, !BEER and !ENGAGE10
LOL An abuse of the rhetorical definition... in this case.
But thanks for all your support and attention mate... I definitely missed many people (and still do) in this almost 4 years of DPoS blockchain. That's the fun about keeping on loving this BC...
Cheers @forykw, Thank You For Inviting @rishi556 To A Glass Of WINE.
At The Moment You Don't Have Enough WINE Staked To Order A Glass Of WINE From @wine.bot.
Please Refill Your Barrels By Directly Buying WINE From Marketplace & Make Sure You Have Atleast 25.00 WINE In Staked Wallet.
Total Purchase : 24809.918 WINE & Last Price : 0.290 HIVE
HURRY UP & GET YOUR SPOT IN WINE INITIAL TOKEN OFFERING -ITO-
WINE Current Market Price : 0.273 HIVE
ENGAGEtokens.View or trade
BEER.Hey @rishi556, here is a little bit of
BEERfrom @forykw for you. Enjoy it!Learn how to earn FREE BEER each day by staking your
BEER.View or trade
BEER.Hey @rishi556, here is a little bit of
BEERfrom @forykw for you. Enjoy it!Learn how to earn FREE BEER each day by staking your
BEER.