Skip to main content

Different types of SAP log replication

There are 4 log replication modes:

 Synchronous in-memory (syncmem): This is the default log replication mode. In this mode, primary node waits for the acknowledgement confirming the log has been received by the secondary node before committing any transactions. 

 Basically, primary node waits until secondary node has received data and as long as the replication status is ACTIVE for all services, there will not be any data loss. 

 The waiting period of the primary system is determined by parameter named logshipping_timeout in global.ini. Its default value is 30 secs and if there is no acknowledgement received from the secondary node after 30 secs, primary node continues without replicating the data.
Synchronous (sync): In this mode, primary node waits for acknowledgement confirming the log has been received AND persisted by the secondary node before committing any transactions.
The key benefit of this option compared to syncmem is the consistency between primary and secondary nodes. You know that the primary node will not commit any transactions until secondary node received and persisted the logs.
Like syncmem, the waiting period of the primary system is also 30 secs by default and determined by logshipping_timeout parameter in global.ini. If there is no acknowledgement received from the secondary node after 30 secs, primary node continues without replicating the data.
Synchronous (full sync): Full sync replication was introduced with SPS08 as an additional option for the sync mode. This mode provides absolute zero data loss no matter what because primary node waits until secondary node received the logs and persisted them on the disk; the transaction processing on the primary node is blocked until secondary system becomes available. This ensures no transaction can be committed on the primary node without shipping the logs to the secondary site.
Asynchronous (async): In this option, primary node does not wait any acknowledgement or confirmation from the secondary node, it commits the transactions when it has been written to the log file of the primary system and sent redo logs to the secondary node asynchronously. 

 Obviously, this option provides the best performance among all four options as the primary node does not have to wait for any data transfer between nodes, or I/O activity on the secondary node. However, async mode is more vulnerable to data loss compared to other options. You might expect some data loss during an (especially unexpected) failovers.
Labels:

Comments

Post a Comment

Popular posts from this blog

Error: could not find function "read.xlsx" while reading .xlsx file in R

Got this during the execution of following command in R > dat Error: could not find function "read.xlsx" Tried following command > install.packages("xlsx", dependencies = TRUE) Installing package into ‘C:/Users/amajumde/Documents/R/win-library/3.2’ (as ‘lib’ is unspecified) also installing the dependencies ‘rJava’, ‘xlsxjars’ trying URL 'https://cran.rstudio.com/bin/windows/contrib/3.2/rJava_0.9-8.zip' Content type 'application/zip' length 766972 bytes (748 KB) downloaded 748 KB trying URL 'https://cran.rstudio.com/bin/windows/contrib/3.2/xlsxjars_0.6.1.zip' Content type 'application/zip' length 9485170 bytes (9.0 MB) downloaded 9.0 MB trying URL 'https://cran.rstudio.com/bin/windows/contrib/3.2/xlsx_0.5.7.zip' Content type 'application/zip' length 400968 bytes (391 KB) downloaded 391 KB package ‘rJava’ successfully unpacked and MD5 sums checked package ‘xlsxjars’ successfully unpacked ...
Post a Comment
Read more

What is Tensor Parallelism and relationship between Buffer and GPU

  Tensor Parallelism in GPU Tensor parallelism is a technique used to distribute the computation of large tensor operations across multiple GPUs or multiple cores within a GPU .   It is an essential method for improving the performance and scalability of deep learning models, particularly when dealing with very large models that cannot fit into the memory of a single GPU. Key Concepts Tensor Operations : Tensors are multidimensional arrays used extensively in deep learning. Common tensor operations include matrix multiplication, convolution, and element-wise operations. Parallelism : Parallelism involves dividing a task into smaller sub-tasks that can be executed simultaneously. This approach leverages the parallel processing capabilities of GPUs to speed up computations. How Tensor Parallelism Works Splitting Tensors : The core idea of tensor parallelism is to split large tensors into smaller chunks that can be processed in parallel. Each chunk is assigned to a different GP...
Post a Comment
Read more

What is the benefit of using Quantization in LLM

Quantization is a technique used in LLMs (Large Language Models) to reduce the memory requirements for storing and training the model parameters. It involves reducing the precision of the model weights from 32-bit floating-point numbers (FP32) to lower precision formats, such as 16-bit floating-point numbers (FP16) or 8-bit integers (INT8). Bottomline: You can use Quantization to reduce the memory footprint off the model during the training. The usage of quantization in LLMs offers several benefits: Memory Reduction: By reducing the precision of the model weights, quantization significantly reduces the memory footprint required to store the parameters. This is particularly important for LLMs, which can have billions or even trillions of parameters. Quantization allows these models to fit within the memory constraints of GPUs or other hardware accelerators. Training Efficiency: Quantization can also improve the training efficiency of LLMs. Lower precision formats require fewer computati...
Post a Comment
Read more