Talk:Distributed Shared Memory - Revision history

Yacoub A.: /* IVY */

2008-10-01T05:00:03Z

IVY

← Older revision		Revision as of 05:00, 1 October 2008
Line 22:		Line 22:
	[[user:adam_k\|Adam_K]]: (Answer) An important characteristic of DSM in general is to make use of processor cycles and memory resources in a loosely coupled system that otherwise would remain idle. This is desirable for more than just speeding up divide and conquer scenarios, but also for creating a capacity of memory larger than any individual system could accomodate. Such an environment is neccessary for certain applications, involving very large models or simulations where it may not be practical to partition the data set.		[[user:adam_k\|Adam_K]]: (Answer) An important characteristic of DSM in general is to make use of processor cycles and memory resources in a loosely coupled system that otherwise would remain idle. This is desirable for more than just speeding up divide and conquer scenarios, but also for creating a capacity of memory larger than any individual system could accomodate. Such an environment is neccessary for certain applications, involving very large models or simulations where it may not be practical to partition the data set.

	(Question) How would several processes trying to access the same data block in the dynamic distributed manager algorithm degrade significantly the performance of the system?		[[user:Yacoub A.\|Yacoub A.]]:(Question) How would several processes trying to access the same data block in the dynamic distributed manager algorithm degrade significantly the performance of the system?

	== Current DSM systems? ==		== Current DSM systems? ==

Yacoub A.: /* IVY */

2008-10-01T04:58:38Z

IVY

← Older revision		Revision as of 04:58, 1 October 2008
Line 21:		Line 21:

	[[user:adam_k\|Adam_K]]: (Answer) An important characteristic of DSM in general is to make use of processor cycles and memory resources in a loosely coupled system that otherwise would remain idle. This is desirable for more than just speeding up divide and conquer scenarios, but also for creating a capacity of memory larger than any individual system could accomodate. Such an environment is neccessary for certain applications, involving very large models or simulations where it may not be practical to partition the data set.		[[user:adam_k\|Adam_K]]: (Answer) An important characteristic of DSM in general is to make use of processor cycles and memory resources in a loosely coupled system that otherwise would remain idle. This is desirable for more than just speeding up divide and conquer scenarios, but also for creating a capacity of memory larger than any individual system could accomodate. Such an environment is neccessary for certain applications, involving very large models or simulations where it may not be practical to partition the data set.

			(Question) How would several processes trying to access the same data block in the dynamic distributed manager algorithm degrade significantly the performance of the system?

	== Current DSM systems? ==		== Current DSM systems? ==

Dave: /* DSM Implementations? */

2008-09-29T18:17:45Z

DSM Implementations?

← Older revision		Revision as of 18:17, 29 September 2008
Line 71:		Line 71:

	[[user:Azalia\|Azalia]] : (Answer) As far as the discussed material in the dissertation and the other paper, I did not notice any discussion about DSM security. However, I did some further research about this and found this paper : (http://www-static.cc.gatech.edu/~milos/rogers_pact06.pdf) This paper has been published in 2006 and according to the paper it is the first work related to DSM security.		[[user:Azalia\|Azalia]] : (Answer) As far as the discussed material in the dissertation and the other paper, I did not notice any discussion about DSM security. However, I did some further research about this and found this paper : (http://www-static.cc.gatech.edu/~milos/rogers_pact06.pdf) This paper has been published in 2006 and according to the paper it is the first work related to DSM security.

			[[user:Dave\|Dave]] : (Answer)After browsing the article mentioned above, it seems as though the earlier systems offered only software security measures, but failed to implement hardware security measures. This became apparent in later years as people began to attach "snooping devices" to various parts of the systems. They could then use these devices to capture information that happened to be flowing through that part of the system. Since the article is from 2006, I assume there are probably a few DSM systems out there that implement some sort of hardware security measures, however it seems to be a relatively new idea.

	== Performance vs. Ease of Use ==		== Performance vs. Ease of Use ==

Laszlo: /* Performance vs. Ease of Use */

2008-09-26T20:50:25Z

Performance vs. Ease of Use

← Older revision		Revision as of 20:50, 26 September 2008
Line 83:		Line 83:

	[[user:adam_k\|Adam_K]]: (Question) Regarding reliability more specifically: how does DSM deal with node failure? That is, if a node with the only copy of a certain block of the shared address space, without explicit redundancy, how does the system gaurd against loss of data?		[[user:adam_k\|Adam_K]]: (Question) Regarding reliability more specifically: how does DSM deal with node failure? That is, if a node with the only copy of a certain block of the shared address space, without explicit redundancy, how does the system gaurd against loss of data?

			[[user:laszlo\|Laszlo]]: (Answer) That is a very good question about reliability of nodes. The advantage of having DSM is the ease of use because accessing memory by address is very easy to do. If a system runs out of memory and has to destroy a page of memory it will kill the process that owned that memory. If a node in the network is lost, one would assume the process would have to be terminated because it is hard to handle the condition of missing memory as a programmer. I think that usually the node with the required memory would also contain that process, so losing that node would implicitly kill the process. Maybe it could keep the process around in case the node came back online. To guard against the loss of data, a DSM system could be set up with a certain number of extra copies of each page. If the physical network was known to be very reliable the redundancy could be removed to make more memory available. This flexibility is one of the advantages of a DSM system.

Laszlo: /* Difference between DSM and NUMA? */

2008-09-26T20:36:23Z

Difference between DSM and NUMA?

← Older revision		Revision as of 20:36, 26 September 2008
Line 55:		Line 55:
	[[User:eltonc\|eltonc]]: From what I have read, the NUMA architechture first tries to see if the data requested can be found in local memory, but since there is a limitation on how much local memory a computer can have, it then tries to access the remote memory (memory on a different node than the CPU currently running the process). This remote memory is usually on the same machine where as with DSM, the memory is usually on a different computer.		[[User:eltonc\|eltonc]]: From what I have read, the NUMA architechture first tries to see if the data requested can be found in local memory, but since there is a limitation on how much local memory a computer can have, it then tries to access the remote memory (memory on a different node than the CPU currently running the process). This remote memory is usually on the same machine where as with DSM, the memory is usually on a different computer.

	[[User:laszlo\|Laszlo]] NUMA and DSM both include separate memory for each processor for fast access, as well as the ability to access memory from other processors' memories. The different is that DSM is distributed and NUMA is not. NUMA does provide a means for using more than one processor to bypass the performance bottleneck, however NUMA was designed for SMP machines, where all the processors are linked together within the same physical machine. The main point of DSM is to have memory shared between physically distinct machines. This provides both the fault tolerance (a single multiprocessor machine will fail all at once), as well as the flexibility to add or remove machines to make the distributed system bigger or smaller. It is discussed in the paper that DSM would allow the extra cycles on each workstation to be made use of when the primary user was not making use of their workstation. NUMA does not have the option to do this, because a single multiprocessor machine cannot be used like separate workstations can. Additionally a DSM system simplifies the use of other system components such as the hard disk. Since each processor has their own peripherals, disks and circuits there may be less problems using these resources.		[[User:laszlo\|Laszlo]]: (Answer) NUMA and DSM both include separate memory for each processor for fast access, as well as the ability to access memory from other processors' memories. The different is that DSM is distributed and NUMA is not. NUMA does provide a means for using more than one processor to bypass the performance bottleneck, however NUMA was designed for SMP machines, where all the processors are linked together within the same physical machine. The main point of DSM is to have memory shared between physically distinct machines. This provides both the fault tolerance (a single multiprocessor machine will fail all at once), as well as the flexibility to add or remove machines to make the distributed system bigger or smaller. It is discussed in the paper that DSM would allow the extra cycles on each workstation to be made use of when the primary user was not making use of their workstation. NUMA does not have the option to do this, because a single multiprocessor machine cannot be used like separate workstations can. Additionally a DSM system simplifies the use of other system components such as the hard disk. Since each processor has their own peripherals, disks and circuits there may be less problems using these resources.

	== DSM Implementations? ==		== DSM Implementations? ==

Bo at 20:35, 26 September 2008

2008-09-26T20:35:57Z

← Older revision		Revision as of 20:35, 26 September 2008
Line 79:		Line 79:

	[[User:Ywahyudi\|Yohan]]: (Question) Speaking about performance and reliability of a system, we know that performance is a huge issue for Li. And on chapter 7.4, he said that he didn't take reliability aspect into consideration at all. However, we all know that a fast application which has an almost linear running time will be useless if that application cannot deliver the expected result. So my question is, is performance a lot more important comparing to reliability? At least during Li's era? and What aspects (during Li's era) made performance of a system more important than its reliability? (Recall that when RPC was first introduced, reliability and security were not addressed too)		[[User:Ywahyudi\|Yohan]]: (Question) Speaking about performance and reliability of a system, we know that performance is a huge issue for Li. And on chapter 7.4, he said that he didn't take reliability aspect into consideration at all. However, we all know that a fast application which has an almost linear running time will be useless if that application cannot deliver the expected result. So my question is, is performance a lot more important comparing to reliability? At least during Li's era? and What aspects (during Li's era) made performance of a system more important than its reliability? (Recall that when RPC was first introduced, reliability and security were not addressed too)

			[[User:BoWang\|Bo]]: (Answer) Performance and reliability more or less go hand in hand. If the RPC implementation loses packets, those packets would need to be resent and the programs would hang until some result was received, which would decrease the performance. Li's main reason for not factoring in reliability is the fact that his experiment set was reproducible such that if one computation failed to return, he would just recompute it.

	[[user:adam_k\|Adam_K]]: (Question) Regarding reliability more specifically: how does DSM deal with node failure? That is, if a node with the only copy of a certain block of the shared address space, without explicit redundancy, how does the system gaurd against loss of data?		[[user:adam_k\|Adam_K]]: (Question) Regarding reliability more specifically: how does DSM deal with node failure? That is, if a node with the only copy of a certain block of the shared address space, without explicit redundancy, how does the system gaurd against loss of data?

Laszlo: /* Difference between DSM and NUMA? */

2008-09-26T20:35:37Z

Difference between DSM and NUMA?

← Older revision		Revision as of 20:35, 26 September 2008
Line 54:		Line 54:

	[[User:eltonc\|eltonc]]: From what I have read, the NUMA architechture first tries to see if the data requested can be found in local memory, but since there is a limitation on how much local memory a computer can have, it then tries to access the remote memory (memory on a different node than the CPU currently running the process). This remote memory is usually on the same machine where as with DSM, the memory is usually on a different computer.		[[User:eltonc\|eltonc]]: From what I have read, the NUMA architechture first tries to see if the data requested can be found in local memory, but since there is a limitation on how much local memory a computer can have, it then tries to access the remote memory (memory on a different node than the CPU currently running the process). This remote memory is usually on the same machine where as with DSM, the memory is usually on a different computer.

			[[User:laszlo\|Laszlo]] NUMA and DSM both include separate memory for each processor for fast access, as well as the ability to access memory from other processors' memories. The different is that DSM is distributed and NUMA is not. NUMA does provide a means for using more than one processor to bypass the performance bottleneck, however NUMA was designed for SMP machines, where all the processors are linked together within the same physical machine. The main point of DSM is to have memory shared between physically distinct machines. This provides both the fault tolerance (a single multiprocessor machine will fail all at once), as well as the flexibility to add or remove machines to make the distributed system bigger or smaller. It is discussed in the paper that DSM would allow the extra cycles on each workstation to be made use of when the primary user was not making use of their workstation. NUMA does not have the option to do this, because a single multiprocessor machine cannot be used like separate workstations can. Additionally a DSM system simplifies the use of other system components such as the hard disk. Since each processor has their own peripherals, disks and circuits there may be less problems using these resources.

	== DSM Implementations? ==		== DSM Implementations? ==

Adam k: /* Performance vs. Ease of Use */

2008-09-26T20:28:28Z

Performance vs. Ease of Use

← Older revision		Revision as of 20:28, 26 September 2008
Line 77:		Line 77:

	[[User:Ywahyudi\|Yohan]]: (Question) Speaking about performance and reliability of a system, we know that performance is a huge issue for Li. And on chapter 7.4, he said that he didn't take reliability aspect into consideration at all. However, we all know that a fast application which has an almost linear running time will be useless if that application cannot deliver the expected result. So my question is, is performance a lot more important comparing to reliability? At least during Li's era? and What aspects (during Li's era) made performance of a system more important than its reliability? (Recall that when RPC was first introduced, reliability and security were not addressed too)		[[User:Ywahyudi\|Yohan]]: (Question) Speaking about performance and reliability of a system, we know that performance is a huge issue for Li. And on chapter 7.4, he said that he didn't take reliability aspect into consideration at all. However, we all know that a fast application which has an almost linear running time will be useless if that application cannot deliver the expected result. So my question is, is performance a lot more important comparing to reliability? At least during Li's era? and What aspects (during Li's era) made performance of a system more important than its reliability? (Recall that when RPC was first introduced, reliability and security were not addressed too)

			[[user:adam_k\|Adam_K]]: (Question) Regarding reliability more specifically: how does DSM deal with node failure? That is, if a node with the only copy of a certain block of the shared address space, without explicit redundancy, how does the system gaurd against loss of data?

Adam k: /* IVY */

2008-09-26T20:25:31Z

IVY

← Older revision		Revision as of 20:25, 26 September 2008
Line 19:		Line 19:

	[[user:RobertB\|RobertB]]:(Answer) While some form of distributed (or hybrid) manager algorithm is the alternative to a centralized one, it too has its own set of issues that need to be considered. With decisions being made in several locations the chances of data corruption, incorrect reads, preemptive writes, etc are all potentially increased. Thus the implementation of a distributed manager would be a lot more complex and time-consuming.. not to mention difficult to debug, which may impact one's decision to go that route.		[[user:RobertB\|RobertB]]:(Answer) While some form of distributed (or hybrid) manager algorithm is the alternative to a centralized one, it too has its own set of issues that need to be considered. With decisions being made in several locations the chances of data corruption, incorrect reads, preemptive writes, etc are all potentially increased. Thus the implementation of a distributed manager would be a lot more complex and time-consuming.. not to mention difficult to debug, which may impact one's decision to go that route.

			[[user:adam_k\|Adam_K]]: (Answer) An important characteristic of DSM in general is to make use of processor cycles and memory resources in a loosely coupled system that otherwise would remain idle. This is desirable for more than just speeding up divide and conquer scenarios, but also for creating a capacity of memory larger than any individual system could accomodate. Such an environment is neccessary for certain applications, involving very large models or simulations where it may not be practical to partition the data set.

	== Current DSM systems? ==		== Current DSM systems? ==

Ywahyudi at 19:40, 26 September 2008

2008-09-26T19:40:41Z

← Older revision		Revision as of 19:40, 26 September 2008
Line 73:		Line 73:

	A semi-related question: did he actually get a PhD for this paper? If so, it speaks to why Yale is not known for computer science.		A semi-related question: did he actually get a PhD for this paper? If so, it speaks to why Yale is not known for computer science.

			[[User:Ywahyudi\|Yohan]]: (Question) Speaking about performance and reliability of a system, we know that performance is a huge issue for Li. And on chapter 7.4, he said that he didn't take reliability aspect into consideration at all. However, we all know that a fast application which has an almost linear running time will be useless if that application cannot deliver the expected result. So my question is, is performance a lot more important comparing to reliability? At least during Li's era? and What aspects (during Li's era) made performance of a system more important than its reliability? (Recall that when RPC was first introduced, reliability and security were not addressed too)

← Older revision		Revision as of 20:35, 26 September 2008
Line 79:		Line 79:

	[[User:Ywahyudi\|Yohan]]: (Question) Speaking about performance and reliability of a system, we know that performance is a huge issue for Li. And on chapter 7.4, he said that he didn't take reliability aspect into consideration at all. However, we all know that a fast application which has an almost linear running time will be useless if that application cannot deliver the expected result. So my question is, is performance a lot more important comparing to reliability? At least during Li's era? and What aspects (during Li's era) made performance of a system more important than its reliability? (Recall that when RPC was first introduced, reliability and security were not addressed too)		[[User:Ywahyudi\|Yohan]]: (Question) Speaking about performance and reliability of a system, we know that performance is a huge issue for Li. And on chapter 7.4, he said that he didn't take reliability aspect into consideration at all. However, we all know that a fast application which has an almost linear running time will be useless if that application cannot deliver the expected result. So my question is, is performance a lot more important comparing to reliability? At least during Li's era? and What aspects (during Li's era) made performance of a system more important than its reliability? (Recall that when RPC was first introduced, reliability and security were not addressed too)

			[[User:BoWang\|Bo]]: (Answer) Performance and reliability more or less go hand in hand. If the RPC implementation loses packets, those packets would need to be resent and the programs would hang until some result was received, which would decrease the performance. Li's main reason for not factoring in reliability is the fact that his experiment set was reproducible such that if one computation failed to return, he would just recompute it.

	[[user:adam_k\|Adam_K]]: (Question) Regarding reliability more specifically: how does DSM deal with node failure? That is, if a node with the only copy of a certain block of the shared address space, without explicit redundancy, how does the system gaurd against loss of data?		[[user:adam_k\|Adam_K]]: (Question) Regarding reliability more specifically: how does DSM deal with node failure? That is, if a node with the only copy of a certain block of the shared address space, without explicit redundancy, how does the system gaurd against loss of data?