All About Circuits’ Post

A RAID (Redundant Array of Independent Disks) system is a #storage technology that combines multiple physical drives into a single logical unit for data redundancy, performance improvement, or both. There are different RAID levels, each offering unique benefits such as data mirroring, striping, or parity. Fnd below picture, transferring much knowledge about #RAID

Check out the Tips & Tricks from our partner NetAlly coming at you every Wednesday! [email protected] (941) 721-7111

Paper of the day: Cache-Efficient Top-k Aggregation over High Cardinality Large Datasets Zippy, a new cache-conscious aggregation framework that leverages the skew in the data distribution to minimize data movements where intermediate results may not fit within the local cache of multi-core processors

REx: Rule Explorer project and Detection Engineering Threat Report (DETR) If you found LoFP from http://lolol.farm useful, this provides significantly more insights via a much more powerful platform (built on the Elastic stack!) CC: @br0k3ns0und #elastic #threatradar #threatintelligence #edr #detection #threat

Failure sucks. It’s also something that happens to every data architect at some point and more frequently in the case of HDDs. What if, however, your instance could lose up to half its drives and without losing any data? At MinIO, our Erasure Coding provides data redundancy, fault tolerance and availability. This article focuses on how to manage failing drives, covering drive recovery and replacement best practices that minimize operational disruption. https://hubs.li/Q025wVVw0

#Protip The Challenges of Packet Captures (3 Min read) #NetAlly #CableAndConnections https://ow.ly/T1Zg50QhZvf

If a memory has dual use, the data sizes for each use case may not match. You can pre-compute the Least Common Multiple of the data sizes to make sure either use case can store an integer number of items. This simplifies addressing and multiplexing. (more on this later...) https://lnkd.in/gNCnRdJq

SIEVE: Cache eviction can be simple, effective, and scalable

Example: Let us understand with an example, Memory Read Request requesters device core/SL sends request to TL with information such as memory address, transaction type, amount of data to read calculated in dwords, traffic class and other attributes. TL uses this information and creates Memory read TLP. Depending on address size(32/64 bit) a 3DW or 4DW header will be created. TL also adds requester ID(bus#, device#, function#), it is required for the completer to send completion. At requester TL, generated TP will be placed in appropriate Virtual Channel Buffer, once flow control logic confirms there is enough space in receiver's virtual channel buffer, MRd TLP will be sent to the DLL. DLL adds sequence number and LCRC, and a copy of TLP with sequence number and L CRC will be stored in replay buffer of DLL and packet will be forwarded to the PL. PL adds start and end characters and the packet undergoes byte stripping across available lanes, scrambled and encoded and then serialized on each lane and will be transmitted differentially across link to the other device. Completer performs the reverse operations, it de-serializes incoming bits, they will be passed through elastic buffer then decoded back from all Lanes, de-scrambled and un-stripped. Start and end characters added by requester will be deleted in completer PL. DL checks for LCRC errors and checks sequence number, if no errors then it creates Ack with same seq.no that’s was used in read request. To ack as discussed, 16 bit CRC will be appended and the created DLLP will sent to PL, PL adds framing packets and transfers Ack DLLP to PL of requester. Requester PL removes framing symbol from Ack DLLP and forwards it to DL, if CRC is valid, the stored TLP with the acknowledged sequence number will be removed from DL replay buffer. If Nak DLLP was received, requester will re-send a copy of stored TLP. Completer's DLL sends TLP to its TL. Completers TL places the TLP in a VC receive buffer. ECRC is optional. If present ECRC check happens and if no errors are found, content of TLP like memory read transaction type, amount of data requested, traffic class will be forwarded to the completer's SL.

In this Cheryl Watson's Tuning Letter reprint, Todd Havekost addresses the most common questions clients have about upgrading to a CPC in a lower speed range. While sub-capacity models may not be a one-size-fits-all solution, this article raises awareness about the growing significance of sub-capacity models in many environments.🌍💡 Read the full article here 👉 https://lnkd.in/g4b6PTeV Cheryl Watson #mainframe #TechInnovation #dataanalytics #ibmstorage