CyberTouch - Enjoy Your Digital Life

Toyota VS Mazda

admin — Thu, 20 Nov 2008 00:56:53 +0000

Toyotas. Period. I am on my 2nd camry since 1992. That’s the same car for 16 years…and counting…this one is a 1999 LE……basic car but you become attached and for some very, very good reasons.

1) Reliability. I have just at 100,000 on this one and so far, I replaced a battery (after 8 years!), two-sets of tires (normal wear and tear), front brakes twice (normal wear and tear), rear brakes once, synthetic oil since day one, some wiper blades, a transmission flush twice, and that’s it.

The power windows, door locks, AC, etc, etc… have never failed me and I live in Atlanta where driving is an art form.

2) Resale. I’ve had a friend offer me 6k last summer for it. I paid 17K plus tax, title, tag when new and never looked back…..had I sold it, I would have paid 11k for 8 plus years of transportation….not counting routine maintenance…..I’ve decided to keep it because frankly, I can’t find a used car this good…anywhere…….I’ve decided that I am going to drive this thing into the ground….and still get 2 or 3k when I’m done!

Smart money says, it will last another 3-5 years before I really have any issues with it and at that point, who can complain? Find me ANY American car, that will perform like that??? They are that good.

The Mazda’s??? Well, take a look at resale and that should tell you what you need to know…..for about $1,500 more, you can get a similar Honda or Toyota, why bother? Why buy misery? While the Mazda’s aren’t bad, I don’t see many 9/10 year old ones out there….

I will say that Toyota is the superior company. While no car lasts forever, my Toyota truck with the 22-RE engine has been running strong since 1993 with only minor maintenance and tune ups at appropriate intervals. It has never left me stranded and I am able to do most easy and intermediate work at home.

Mazdas, on the other hand, are not so reliable. Many people don’t know that Mazda is an associate company of Ford. Mazda’s quality assurance is not as good at Toyota’s and you won’t see nearly as many older Mazdas on the road as you will Toyotas. I think that is a good indicator of quality. I see plenty of older Toyotas (even older than mine) on the road and running strong.

Good luck with your choice!

META tags Web content writing norms points

admin — Wed, 05 Nov 2008 14:08:59 +0000

Observation of the new online marketing network marketing seminars on the design of the META tags on a stress analysis of the contents of the Web Design META tags written in the common errors and the consequences of the current site of most of the content of the META tags are writing or the existence of such problems, this is not That the content of the META tags writing how difficult it usually is not enough attention, there is no intention to do it.

What kind of META tag is a reasonable design? To put it simply, is the META tag in the correct description of the main page description of the contents of the summary information on the contents of the summary page and the page containing the core of the keywords, META tag keywords in the page that further the core of the words (of these words Appear on the same page description of the message), if more than one core keywords, separated by commas between keywords.

The following is the author of this lecture Feng Jian page summary of the contents of the META tags writing norms points.
(1) on the keywords in META tags (keywords) on the design features of

Select pages with content most relevant to the core keywords, and the number of words without too much, but there is no need and a pile of a large number of words;
Key words in the keyword should appear in the same description of the contents;
Different comma between the words (English punctuation) separated.
(2) on the META tag in the page description (description) of the design features of the

Page described as natural language rather than keywords list (keywords and design just the opposite);
As far as possible, an accurate description of the core content of the page, usually page summary of the contents of the message, that is, in the hope that the search engine search results to display a summary of information;
Description of the page contains effective keywords;
Description of the contents of the page with website content heading a high degree of correlation;
Description of the contents of the web page with the main contents of a high degree of correlation;
Described in the pages do not have too many words, no more than general search engines to retrieve information from the summary results of the most number of words (usually 100 characters, a slight difference between the different search engines).

More about some of the norms of writing META tags, I have been a number of sites verified, it can achieve very good results. But to show that these META tags writing norms points only the general rules do not necessarily suitable for all pages (for example, to roll over some of the content of the page to update the design of how the META tags to describe information in order to reflect the changing content? - For the time being here does not give details, interested readers to think about, will observed the new on-line marketing at the appropriate time devoted to this particular problem), not necessarily for all search engines are effective, with the search engine search Changes in rules, the META tag writing method also requires a flexible application-specific situation and making necessary adjustments, the only fan to regulate the network marketing search engine optimization as a reference.

Obama being involved in Jerimiah Wright’s church

admin — Wed, 05 Nov 2008 14:03:34 +0000

Because white people feel the need to cater to black people because they still feel they owe them something from what happened back in the day. Time to get over it. Time for black people to GET OVER IT.

I pray to God that black people now stop pulling the race card every chance they get now that Obama is in office. However, I think it’ll just make things worse since they already think they can get a free ride and special treatment because the most powerful man in the country is now black.

Just set ourselves back 65 years.

what have you seen of Rev . Right other than 15 to 30 seconds that Hannity plays over and over and over. You truly know nothing about him or his church.

I am not even sticking up for the guy, I am just pointing out that you know little or nothing about him, yet you feel you can judge the man on very little information.

The KKK has raped, beaten, murdered and terrorized black people for over a century. Comparing their actions to the actions of Obama’s FORMER pastor or anyone in his church is misguided at best.

I always find it preferable to think about what I’m saying before I say it and to be fully educated on a topic before commenting. Otherwise someone might say I’m all mouth and no brain and that would be embarrassing because no one has ANY respect for a person who is all mouth and no brain.

“If we love our country, we must love our countrymen.”

– Ronald Reagan

Detailed pre-reading algorithm

admin — Mon, 20 Oct 2008 17:24:49 +0000

Why do I need to read?

1. I / O delay

2. the merger hidden

Second Reading of the pre-preliminary design of the algorithm

Read system calls for the realization of the core functions for do_generic_read. If we do not consider pre-reading and visual sense, using their pseudo-code that should be in the form of the following:

do_genric_read ()
(
for (read system calls need to read all the pages) (
if (page is not pagecache) (
Distribution page by adding pagecache;
Locked page;
Start page of the I / O transmission;
)
At present, waiting for the page I / O completion of the transfer;
copy of the current page of data to the user space;
)
)

In the above, we will streamline the core processes: only read when they start a new page in the I / O; if the page has been pagecache years, or that it contains valid data, either by adding a new page in the pagecache, I have to start / O. Without taking into account I / O error, to simplify this is feasible.

If you consider the pre-reading, there are two forms of pre-reading: The pre-time synchronization with asynchronous read-ahead.
Each system calls need to read to read all the pages in a row, and they should be a one-time pre-synchronization time. “Synchronization” because the need to read these pages to wait for synchronous I / O to complete. When read the previous read in the paper’s position in the order, it is to read the order paper at this time should be asynchronous read-ahead: The start of the read “the last page after” a group of consecutive pages of pre-reading, To read the next in the order ahead of time to prepare the data to form the pipeline pre-reading. “Asynchronous” because the follow-up to read the I / O and data processing is the start of induction.

Special friendship to remind: after the “read” all means read system call.

Detection of the order of reading a special case. When the file is opened after the first reading, and reading the paper is the first time, we have a presumption of goodwill, the follow-up order will be read, and therefore the need for asynchronous read-ahead. If not satisfied with the order of the above-mentioned conditions, to determine a random time. Any random time will be sequences in the order to terminate the current, so the termination of pre-reading act.

Both the use of pre-reading, to the greatest extent possible hidden I / O delay. With the realization of its pseudo-code that into the form below:

do_genric_read ()
(
generic_read_ahead (read need to read all the pages) / * * time to start pre-sync /
if (read order)
/ * Start asynchronous read-ahead * /
generic_read_ahead (read the last page after page of a group of straight);
/ * Wait for the completion of the simultaneous pre-reading * /
for (read need to read all the pages) (
At present, waiting for the page I / O completion of the transfer;
copy of the current page of data to the user space;
)
)

Attention: do_genric_read do not need to wait for pre-reading asynchronous I / O completion of the transfer.

generic_read_ahead (pre-reading all the pages)
(
for (pre-reading all the pages) (
if (page is not pagecache) (
Distribution page by adding pagecache;
Locked page;
Start page of data I / O transmission;
)
)
)

Attention: generic_read_ahead only start page I / O transmission, but it does not wait for the I / O to complete.
Asynchronous read-ahead “for a page” in the paper read at the beginning of an initial value can be set up, such as “read need to read all the pages of” 2 times. In order of time, increasing the amount available (for example, double), until a certain limit, also referred to the back.

To achieve the above-mentioned features are:
1 pre-synchronization time of the page will certainly be used, not wasted; asynchronous read-ahead pages may not be used to be, if not the follow-up time of the order, it may be some or all of the waste.
2 simultaneous pre-reading for current services, read, it is a smaller group of pre-reading; and asynchronous read-ahead is to read the order form line services, there should be more pre-reading group.
3, we start the beginning of all possible pre-reading, which is a “pre-determined time to time”.

Third, to achieve the above-mentioned problems

We use pre-reading in the hope that the procedures in dealing with a group of data to the hard disk in the background to the next batch of data to the CPU is ready to advance, in order to be able to hard disk and CPU pipeline operations. Pre-reading of the pipeline to reach the ideal state is: when to read “Pre-time synchronization,” the page request, the page has been read by the previous “asynchronous read-ahead” reading into the memory, so read without having to wait until I / O transmission .

To consider this situation:

The assumption that read “simultaneous pre-reading” locked on the first page. If it is a pre-read (or more before the read) the “asynchronous read-ahead” to start I / O transmission, and that I / O transmission has not yet been completed. It is clear that the ideal comparison, I / O transmission speeds to be slower than the speed of the process. This may be the I / O system, the burden too heavy, or it may be too fast to deal with the process. In either case, at this time and then start the other pre-reading I / O (with a pre-read issued by the I / O request is unlikely the merger) is not only ineffective, is likely to increase the I / O burden, making I / O Transmission would be delayed. However, implementation, read the simultaneous pre-conditions is the time to start page for all the I / O transmission (if it was not read the previous asynchronous read-ahead to start any) and may lead to performance problems.

Consider this situation:

Assuming the order is read, “asynchronous read-ahead” on the first page is not locked on I / O transfer has been completed. If this page is from “a pre-read” time to start pre-asynchronous I / O, read-ahead for the pipeline to reach the ideal state, when the next round of the pre-start time is feasible. However, if this page is from “read previous prior to read” Start I / O, obviously, I / O transmission speed in the process of running (the process of calculating the amount may be too much), pre-reading pages of a sufficient number of relatively very At this time, other pre-boot time I / O, will be made in advance of the page read too much too soon to occupy space pagecache, and follow-up time once the order is not time, lead to I / O processing time and space pagecache waste. However, to achieve, to read as long as the order of time, the “asynchronous read-ahead” is unconditional and will lead to performance problems.

Four new designs

The combination of the above analysis, pre-reading to maximize the benefits must be pre-read the page to take into account two factors: its I / O has been sent by the previous read or read before the start; it is still in I / O transmission. The former record of the need to “pre-reading of history”, which need to track the state of locking the page. Records of the “pre-reading of history” can clearly know what the new pre-reading where to start. According to two factors need to read a combination of characteristics, to decide whether to start pre-reading of I / O. We consider pre-reading pre-recorded reading of history, and then the “pre-determined time to time” into a “dynamic time for pre-reading”: In dealing with each page read operation, the implementation of the “state of the relevant pages of this” pre-reading: From the page based on pre-reading and pre-locked state of reading history, dynamic decide whether to activate the pre-reading I / O. The basic principle is: When the page is from “In the past, pre-reading” Start I / O and locked when not to launch a new pre-reading; when the page is from “the latest” pre-start time I / O, and Page did not lock Should launch a new pre-reading. It should be emphasized that: “In the past, pre-reading,” including not only “the most recent pre-reading,” but also includes the earlier pre-reading.

We should also be noted that when there is a pre-reading pages and bind together, we can not achieve the same as the original, independent start against Reading in the order of asynchronous read-ahead, and pre-reading function can not determine the order of their time. Therefore, it should be pre-reading function to send a sign that read in the order. Pseudo-code is as follows:

do_genric_read ()
(
int reada_ok = 0; / * whether that sequence of signs reading * /
if (read order) reada_ok = 1;
for (read need to read all the pages) (
if (page is not pagecache) (
Distribution page by adding pagecache;
Locked page;
Start page of the I / O transmission;
)
generic_read_ahead (page, reada_ok, file); / * this page and state-related pre-reading, page refers to the current page * /
At present, waiting for the page I / O completion of the transfer;
copy of the current page of data to the user space;
)
)

generic_read_ahead (page, reada_ok, file)
(
According to the reada_ok logo, page locking and the state of pre-reading of history, to determine in advance whether or not to start reading I / O;
)

Five detailed analysis of the design generic_read_ahead

From the visual point of view, we hope to achieve in the cycle to read these pages, a reader will trigger a simultaneous pre-reading and another reading would trigger asynchronous read-ahead, the other will not trigger pre-reading time. In this way, with the realization of the original semantically equivalent. (The real situation is not necessarily the case, the back will see examples)

1 pre-recorded reading of history

Given to pre-recorded reading of history. We call pre-reading the last page of a collection of pre-reading group, it is clear that pre-recorded reading group is necessary, we will record pre-reading group in the object file. In accordance with previous analysis, when read in the order hit the pre-reading group, if the page does not lock, is to promote pre-reading the next best time: the need to launch a new round of pre-reading of I / O, closely follow the pre-reading group in the rear of the Location of the new pre-reading group, to include pre-reading of the new page. If the page has been locked, not to promote pre-reading. As a result, each of the follow-up to the time the order may have a new pre-reading group, then the need for all of these pre-recorded reading of this group? We have taken note of that trigger the establishment of a new pre-reading group read, read the first page contains a pre-reading group, the follow-up to the time the order may still visit a pre-reading group, but can not visit a Group pre-reading before the page. As a result, the order of the time, just record the last two pre-reading group can be. We will read the last two pre-mix and pre-reading into a window, as a general pre-reading in the history of the object file. Of course, if only reads the front of a pre-read, a read-only group, while pre-time window is pre-reading group. Clearly, the time when the pre-move forward, a new set of pre-reading group, the pre-window time is also simultaneously. It should be emphasized that because of the need to distinguish between the most recent pre-reading and pre-reading before, so pre-reading window not a perfect substitute for pre-reading group. Group pre-reading and pre-set time window, and each is closely related to the size of the read-ahead, we will “advance the next time the size of the” field f_ramax also recorded in the file object.

2 pre-start time to time

With pre-reading group with pre-defined time window, on the pre-start time to read the basic principles can be described as preliminary: When reading the page and have been Lock hit the window of time in advance, not to launch a new pre-reading; when read Page did not hit the lock and pre-reading group, it should launch a new pre-reading. Under this principle, bound, read in the cycle to read the page, we do not know when to start a pre-reading so that it corresponds to the “pre-synchronization time”? When the time is in the order, which reada_ok for 1:00, when another pre-start time, so that it corresponds to the “asynchronous read-ahead”?

To consider this. Read in the pages to read, part of the former page in the window of time in advance, are waiting for I / O completion of the transfer. In the pre-inspection tour outside the window, reading the first page, as it had not been in front of the pre-start time I / O, is likely to pagecache not until the page read operation, will join the pagecache page and start the I / O, this When the page was locked, the need to wait for the I / O operation to complete. And then need to read the page and, of course, are not likely in the middle pagechache. Therefore, at this time is to start “pre-synchronization time,” the best time. It should be noted that, when read in the pages of the window, but did not lock pages, pages that already in pagecache (such as other memory into the process) at this time, no need to start pre-synchronization time, because do not have to wait for the I / O Completion of the operation, “simultaneous pre-start time,” the time will not be hidden.

Launched simultaneous pre-reading, will set up a new pre-reading groups, including the subsequent need to read all the pages. In the forward to continue to read to read the page, a page not found lock (this page is bound in the pre-reading group), if read is the order of reading, this time, should trigger a new pre-reading, that is, “asynchronous read-ahead.” Of course, if not read the order of reading, it should not trigger asynchronous read-ahead.

As a result, pre-reading time to start the basic principles can be summed up as follows:
When a page is not locked and read-ahead window, the pre-start time synchronization.
b When the time is the order of (reada_ok 1), pages not locked, and in the pages of pre-reading group, asynchronous read-ahead to start.
c When the time is not in the order, prohibition of asynchronous read-ahead.
d other circumstances, can not be pre-start time.

The need to mention is that in order to wait for the completion of the simultaneous pre-reading, asynchronous read-ahead to start as soon as possible and completed, making the order of reading between the highest degree of water, asynchronous read-ahead needed to activate low-level I / O device driver Procedures to ensure that a new page to be read as soon as possible. But noted that too often low-level activation of I / O device drivers, will affect the I / O request of the merger, the system is detrimental to performance.

3 pre-reading process

Now to discuss pre-reading and pre-reading group set up the window and push forward the process. When read the first time to open the file, pre-reading and pre-initialization time window is empty, saying that “the next time the size of the pre-” f_ramax initial value of 0. We have taken note of that, read the first time to read documents, pre-reading window is empty, it is likely to read the page pagecache not, in and start pagecache added to the I / O before locking. In order to support the back will trigger the simultaneous pre-reading, set for f _ramax need to read to read the pages. Once synchronization has been triggered by pre-reading, will start the follow-up f_ramax page of the I / O transmission, and set up pre-reading group, including all pre-reading the page, pre-reading and pre-window reading groups to keep consistency.

Then, to make f_ramax double for the order after the time to prepare. Since the time of the order, we need to consider the appropriate size of the pre-reading: The pre-read size is too small, there should be less than the effect of performance; too much time in advance, and may be included in the program do not need too many pages , To start a large number of invalid I / O, a waste of resources. To this end, a 2-fold expansion of the process: f_ramax double in each pre-reading, the next time the order will double the pre-reading the page. When time is of the order, the expansion process will be repeated, f_ramax will be doubled each time until the system to achieve the set pieces of equipment on the limit.

When an order is read, read the page in the cycle before f_ramax to determine whether or not to be less than read a few pages, if so, will f_ramax set to the latter. The logic here is: as long as they can read, whether synchronous or asynchronous, it should at least read the first pre-reading needs of all the pages. Read the assumption of a page, triggering pre-reading and follow-up to the start page f_ramax the I / O transmission, including the need to read all the pages. We have set up pre-reading group, including all pre-reading the page, read-ahead window for the last two pre-reading group and redouble f_ramax.

When a random read. Because it is time for pre-order service time, Pre-reservation read a random reading of history makes no sense at all. We time remove pre-history: pre-reading and pre-purchase time window is empty, f_ramax initial clearance 0. At this point, the file with the state to open the file to read for the first time the state is similar. To set up f_ramax We also need to read the page, the back for pre-preparation time.

4 order to relax standards

Read the current requirements of the order of the previous read in the paper’s position in the order, and its strict definition of the two positions should be closely linked. With the pre-window time, we may be able to relax standards in the order: as long as readers read the first page of hits pre-window time (read a pre-set), that the order is read. That would allow the read-ahead algorithm even more radical, pre-reading may be better.

Six simultaneous pre-reading and pre-induction time of the actual situation of each other

In order to deepen their understanding of pre-reading method, we combined the actual scenes described in sync with the pre-reading asynchronous read-ahead.

Random Reading 1

If you read a random, asynchronous read-ahead will be prohibited. In general, it read all the pages, will trigger a simultaneous pre-reading operation. In extreme cases, such as read to read all the pages have been “other process” into memory and therefore did not lock, the first time in advance will not trigger. However, this need not pre-read.

Reading the order of 2

When this is read in the order of the need to start asynchronous read-ahead.

Reading the order of some combination of circumstances
not a pre-reading
Pre-reading time window within the page, the page has been locked, with no corresponding pre-reading. Pre-reading time outside the window of the page, the page did not lock, there is no pre-induction time. As a result, a pre-reading did not occur. This should be a very extreme case of coincidence.

b an asynchronous read-ahead
Assumptions needed to read to read the page did not exceed pre-reading window.
First time in the pre-window reading, but not in pre-reading group pages, regardless of whether the lock would not carry out pre-reading. Pre-reading and then read the group’s page, which has locked the page will not lead to simultaneous pre-reading, a page not found at this time locks, trigger asynchronous read-ahead: The start page related to the I / O data transfer, including the need to read All pages, and set a new pre-reading and pre-window reading group. Under the new pre-reading group does not include the need to read the page, read read “follow-up page,” regardless of whether the page lock, will not lead to pre-reading.

c twice asynchronous read-ahead
Assumptions needed to read the page to read more than the pre-window time.

Asynchronous read-ahead for the first time the case above, but in the new pre-reading group includes the need to read all the pages. So start with asynchronous read-ahead for the first time a similar situation: first read in the new pre-window reading, but not in the new pre-reading group of pages, will not trigger pre-reading. After further reading of the new pre-reading group page, the page will not lead to lock pre-synchronization time. At this point, a page not found locked, so once again trigger a new asynchronous read-ahead, and once again set a new pre-reading and pre-window reading group. At this point, a new pre-reading group can not be read by the need to contain the page. Therefore, read the follow-up time of the operation will not lead to a new pre-reading.

d a simultaneous pre-reading
Assumptions needed to read the page to read more than the pre-window time.
Reading pre-reading group pages, the pages are locked. Pre-read pages outside the window reading, which did not lock the page will not lead to asynchronous read-ahead. At this point, was found locked in a page, so to carry out simultaneous pre-reading: start page related to the I / O data transfer, including the need to read all the pages, and set a new pre-reading and pre-window reading group. Then read the new pre-time window (with pre-agreed time) in the pages of all the pages are locked, so there will be no pre-reading operation.

a pre-read e Synchronous coupled with a pre-induction time
Assumptions needed to read the pages over the pre-window time.
Pre-time synchronization for the first time the case Ibid. I read in a new window of time in advance (and pre-agreed time) in the page, a page not found locked, so once again a new asynchronous read-ahead, and once again set a new pre-reading and pre-window reading group. At this point, a new pre-reading group can not be read by the need to contain the page, read the follow-up time of the operation will not lead to a new pre-reading.

As a result, if read as a sequence of reading and read all the pages in its process, lead to a maximum of two pre-reading. Also noted that, after reading a pre-induction, read by the needs of all the pages are read-ahead in the new window, so then there will not be simultaneous occurrence of pre-reading. In other words, there will not be a “one-time pre-induction coupled with a simultaneous read-ahead” from happening.

Oracle 10G RAC under the OCR and Voting disk management

admin — Mon, 20 Oct 2008 16:57:00 +0000

Voting disk (disk vote on the side):
Voting disk record of the members of the nodes of information, such as the node which includes members of the node to add the information to delete records for the size of 20MB
View voting disk location: crsctl query css votedisk
$ Crsctl query css votedisk
0. 0 / dev/rhdisk3
If the CRS installation process fails, you need to install the need to re-initialize voting disk set, DD can be used or re-roll
dd if = / dev / zero of = / dev/rhdisk3 bs = 8192 count = 2560
Backup votedisk: dd if = / dev/rhdisk3 of = / tmp / votedisk.bak
To restore votedisk: dd if = / tmp / votedisk.bak of = / dev/rhdisk3
Add voting disk mirroring disk:
crsctl add css votedisk / dev/rhdisk7-force
Remove the voting disk mirroring disk
crsctl delete css votedisk / dev/rhdisk7-force

OCR:
OCR records of members of the node configuration information, such as [u] database, ASM, instance, listener, VIP, and other resources of the CRS configuration information can be stored in the device or bare cluster file system, the recommended setting for the size of 100MB
As a way to RAW, a division of RAW, for example: / dev/rhdisk2
If the CRS installation process fails, you need to install the need to re-initialize OCR disk (RAW mode), DD can be used or re-roll
dd if = / dev / zero of = / dev/rhdisk2 bs = 8192 count = 12800
[u] Oracle [/ u] every four hours to launch automatically [u] backup [/ u], and the preservation of the three versions, but there is only one node on
$ Ocrconfig-showbackup
poaisdb1 2008/04/10 22:53:29 / u01/app/oracle/product/10.2.0/crs/cdata/crs
poaisdb1 2008/04/10 18:53:28 / u01/app/oracle/product/10.2.0/crs/cdata/crs
poaisdb1 2008/04/10 14:53:27 / u01/app/oracle/product/10.2.0/crs/cdata/crs
poaisdb1 2008/04/09 06:53:20 / u01/app/oracle/product/10.2.0/crs/cdata/crs
poaisdb2 2008/04/03 03:31:28 / u01/app/oracle/product/10.2.0/crs/cdata/crs
Resumption of OCR: ocrconfig-restore / u01/app/oracle/product/10.2.0/crs/cdata/crs/backup01.ocr
OCR manually export: ocrconfig-export / tmp / ocr_bak
OCR manually import: ocrconfig-import / tmp / ocr_bak

OCR add disk image:
1. Crsctl stop crs stopped by CRS services
2. To create a RAW image of the OCR devices, such as: / dev/rhdisk6
3. Ocrconfig-export with export OCR information
4. Edit / etc / oracle / ocr.loc file, add the line ocrmirrorconfig_loc
$ Cat ocr.loc
ocrconfig_loc = / dev/rhdisk2
ocrmirrorconfig_loc = / dev/rhdisk6
local_only = FALSE
5. Ocrconfig-import with import OCR information
6. Ocr set up check information
$ Ocrcheck
Status of Oracle Cluster Registry is as follows:
Version: 2
Total space (kbytes): 103724
Used space (kbytes): 3824
Available space (kbytes): 99900
ID: 1086971606
Device / File Name: / dev/rhdisk2 Device / File integrity check succeeded
Device / File Name: / dev/rhdisk6 Device / File integrity check succeeded
Cluster registry integrity check succeeded
7. Crsctl start crs last start with CRS services

To add that if the general votedisk is odd, votedisk if the number is less than 50% (equivalent to 50% of firms do not), then the cluster will not be able to form.

Tomcat and JBoss Web distinction

admin — Mon, 20 Oct 2008 16:45:10 +0000

Web2.0 in the wave of technology and a variety of pages are emerging framework for server-side infrastructure has set a higher stability and scalability. In recent years, the open-source middleware as a global leader, JBoss Application Server J2EE in the field of development has become the fastest application server. In market share and service satisfaction on a great success, in no way inferior to other non-open-source competitors, such as WebSphere, WebLogic, Application Server. JBoss Web’s many superior performance is its widely popular.

Tomcat-based core, is better than blue-green

Tomcat server is a free open source Web application servers, advanced technology, performance and stability, and free of charge, so fans of the popular Java and a favorite part of the recognition of software developers. Its run-time occupation of the small system resources, scalability, and load balancing support and services such as e-mail system development and application of commonly used functions. As a small, lightweight application server, Tomcat in the small and medium-sized systems and concurrent users not to visit the many occasions have been widely used, become more popular Web application servers.

JBoss Web and the industry’s best open-source Java Web engines, download the Java community in the largest, most users, the most comprehensive standards support of the Tomcat as part of its core container Servlet engine, and audit and tuning. Tomcat pure performance is limited, in many places there is a lack of performance, such as support for the activities connected, static content, such as documents and HTTPS. In addition to performance problems, Tomcat, another big drawback is that it is a subject of the integration platform, only to run Java applications. When used in Tomcat, often at the same time needed to deploy Apache Web Server to integrate with them. This configuration is more complicated and can not guarantee the performance of the superiority.

Tomcat in JBoss on the basis of their localization, the Tomcat will be embedded in the way of integration into JBoss. JBoss Web by using the APR and Tomcat local hybrid model of technology to solve a lot less than the Tomcat. Mixed model technology from the operating system up-to-date technology to provide the best thread and deal with the incident. The results, JBoss Web to reach the scalability, performance parameters to match or exceed the local Apache HTTP server or IIS. For example, JBoss Web can provide the database connection pool, not only in support of JSP technology such as Java, as well as other Web technologies to support integration, such as PHP,. NET two camps.

Standardization is to reduce the risk of reliance on technology, the best way to protect investors. JBoss Web the first to support the full range of JEE Web standards, a fundamental guarantee the application of “a development, run anywhere” features, so that the finished product can be easily applied in the JBoss Web and other Java Web server move easily between.

Set in a multi-function, high performance

As a Web application servers in the star product, JBoss Web server as a multi-functions. The key features include: full support for Java EE, a high level of scalability, fast static content processing, cluster, OpenSSL, URL rewrite and comprehensive.

JBoss Web server with native features and powerful scalability, support for a wide range of Java-based server is not content to deal with technology, running at the same time JSP, Servlet, Microsoft. NET, PHP and CGI, to provide a single, high-performance The deployment of enterprise-class platform.

Tomcat and compared, JBoss Web resources in the static performance aspects of the visit. JBoss Web components to support the two models - pure Java and Native I / O. Native components in support of the dynamic run rate will remain unchanged, while the static visit to the use of the resources provided by the operating system itself 0 send copies, CPU to reduce consumption, shorten the response time, greatly increase the throughput, mixed-mode support for the connection To achieve the maximum 10,000 concurrent client access at the same time, and the Apache Web server fairly. In the deployment of high-performance operating system, JBoss Web can be used for pure Java and Native I / O support for the two models, making application development time can be quickly cross-platform migration, and in the deployment of high-performance operating system-related environmental Native . As the JBoss Web has solved the static problem of resource access performance, the solution may be to direct it as a powerful distribution LVS object, RHEL and load balancing system, the formation of linear theory, an unlimited extension of the load balancing scene.

OpenSSL is the industry’s most rapid and safe source of transmission components, can take advantage of the operating system and hardware features to achieve safe and efficient loading. JBoss Web integration of OpenSSL, to provide highly effective security service delivery, making security more on stage. Research shows that, JBoss Web in the SSL performance than pure Tomcat four times faster.

URL re-writing function can shorten the URL, to hide the actual path to improve security, and easy-to-user memory type, and the search engine included. Tomcat does not have the URL rewrite feature, JBoss Web can provide a flexible URL rewriting engine to operate in support of an unlimited number of rules and conditions of the rules. URL rewriting can be left to support the URL error handling, or server from time to time have to deal with other issues.

JBoss Web can be run separately, can also be seamlessly embedded in JBoss Application Server, JBoss middleware platform to become a part of. Not only to call back service performance will be enhanced, but also to make use of the JBoss platform to enhance the characteristics of Web applications:

JGroups based on a variety of programs to support the cluster

Arjuna based on the technology of the JTA and JTS transaction processing support

Optimization of the thread pool and the pool of support

JMX console based on the basic management and support for the JBoss On the senior management to maintain support

JBoss AOP-based technology-oriented aspects of the support structure

Hibernate support services component

Professional support team

Most of the industry’s open-source products in innovative technology, but the sustainability of the product life cycle planning, quality assurance, as well as the lack of effective protection for software integrators and end-user criticism. Red Hat to exercise the “professionalization of open source technology” is the perfect solution to this problem.

From the open source community of JBoss Web, specialized in Red Hat’s open source temper, in terms of performance, scalability, stability, security and so on, has become an enterprise-wide reach, and even carrier-class standards of product quality. Red Hat is not only full-time technical team into the JBoss Web development, but also have specialized products for the QA team for quality assurance. Well-integrated testing and compatibility testing to ensure the stability of the JBoss Web itself, and to ensure that its backward compatible with JBoss and other collaboration products, good interoperability.

In the security service system, JBoss opened up to products provided by the experts at the expert level support for open source software as a service backed by a strong ecological model of software. JBoss Inc., as well as a large network of authorized service partners for JBoss Web, including the entire JEMS product suite to provide comprehensive support services. Tomcat and compared, JBoss Web can provide relocation services and expert on-site services in the relocation of services, the application of the guidance of experts from the Tomcat to JBoss Web migration, save time and effort. The unique service subscription model, the full protection of the life cycle of software to enable companies sit back and relax.

Wildlife Conservation

admin — Sun, 05 Oct 2008 13:38:28 +0000

INTRODUCTION
- Conservation methods might include:
- Setting up ‘action plans’ for individual species – these are known as BAPs (Biodiversity Action Plan)
- Establishing National Parks, nature reserves, and similar protected areas.
- Restricting urban and industrial development.
- Reclaiming derelict sites.
- Legally protecting species and habitats.
- Preventing the release of non-native species.
- Controlling pollution.
- Recycling and reusing.
- Only buying environmentally friendly products.
- Limit the exploitation of non-renewable resources.
- Restrict the trade in endangered species.
- Breeding programs in Zoos
- Rare breeds centers.
- Sperm and seed banks.
- Education.

IMPORTANCE
The fact is that no species has ever had such wholesale control over everything on earth, living or dead, as we now have. That lays upon us, an awesome responsibility. In out hands now lies not only our own future, but that of all living creatures with whom we share the earth. When a species leaves, the environment changes. Maybe eventually it changes so much that we humans have to leave as well. Animal and environment problems become people problems.

RESULTS
As a result of conservation, it will protect the existing biodiversity and help us to have a well balanced environment, as each species depend on each other.

Humans horrible beings

admin — Sun, 05 Oct 2008 13:32:35 +0000

I believe that there is more horror and hate in the world then there is love.

For every loving, caring, selfless person, there’s at least five perverts, thieves, self-absorbed dimwits, murderers, or other crooks.

Life on the streets shows you how people REALLY feel about their fellow man. Ever hear that quote, “You can judge a man by how he treats his inferiors”? (Not an exact quote there, but I’m too lazy to look it up). When you live on the street, people flat-out ignore you, or give you dirty looks when you ask for money or food. Cops kick you out of warm sleeping places in the winter. People laughed and pointed and called me a prostitute or a disrespectful welfare brat. (No, I was not a prostitute, nor a bratty teenager who ran away from home because Mommy took away the cell phone).

Some people, fortunately, are lucky to find the loving people in the world. I am EXTREMELY lucky to live in Canada, where they have shelters, welfare money, and free schools. I’d be dead by now if I lived in a third-world country. It’s horrible to think that so many people DO live in third-world countries.

America spends billions on making bombs while children are starving. Celebrities on TV spend hundreds of thousands of dollars on shoes, while there are people in the world with no clean water to drink.

If the good in the world outweighed the evil…this wouldn’t happen. Neither would war, murder, or other crimes. There would be food, water, and shelter for all.

Starving yourself is unhealthy

admin — Sun, 05 Oct 2008 13:23:37 +0000

so few times a day your stomach produces acids to digest your food.

When there is no food in your stomach and your body requires nutriments, the stomach still produces acids and it digests not your food, since there is none, but your stomach, and that’s how people get ulcer, from malnutrition. So if you starve yourself, your stomach will shrink and there is a high chance that you will get an ulcer. Which is a bad thing to have since you can’t heal that. It’s a hole in your stomach that hurts a lot, like hell.

So eat healthy and don’t starve yourself.

starving yourself isn’t healthy. Do you mean that if you done it for a few days so the fat would have to be used therefore you would lose weight? If that is what you mean, if you done that, your stomach would shrink and then when you tried to start eating again you wouldn’t be able to eat much as your stomach would be too small.

Either way DO NOT starve yourself, it is really bad for your body imagine all your organs needing food to work properly and you are depriving them of it. Your body is like a car, you feed it and it works. Your body can’t run on empty.

When your body goes into starvation mode it will hold on to everything, even fat. But if you are really starved, then you are not getting any nourishment and your muscles become emaciated. If you get the nutrients that you need, by definition you are not starving, in fact that is one definition of good health.

Mike Sheahan’s top 50

admin — Sun, 05 Oct 2008 13:16:50 +0000

Michael Osbourne at 48 is crazy. If he played for Freo and had the same season personally, he wouldn’t even be considered for top 200. It’s just because he plays for the Premiers.
Davis did not have a better season than Swan. Davis didn’t have a very good start to the year and Swan was much more consistent.
No way was Tom Harley the 3rd best defender all year. Let alone the 20th best performed all year. He is a key defender who does not play on key forwards - Mackie and Scarlett do that.
Corey at 3. He was good, but not THAT good.
Jarrad McVeigh, Rob Murphy, Kirk and Delideo behind Davis is a joke.
It’s funny how Goodes didn’t even finish in the top 10 for the Swans B&F nor is he on this list and he finished top 5 in the Brownlow, just shows how much the umpires love him.
Brownlow medalist at 9 behind Fevola is a joke.
A lot of it is a joke really and is very Victorian based. Just look at J McVeigh down so low.

Mike Sheahan is the most biased football journalist I have ever encountered.
He allegedly played a few games for Werribee some 120 years ago and couldn’t progress and now has a huge chip on shoulder.
He hates the Footscray/Western Bulldogs.
In his columns in the week of the Brownlow medal he bagged Adam Cooney in saying the Ablett should have won. He bagged Scott West a 7 time B&F at Doggieland and then slammed Dougie Hawkins.
It’s a wonder he didn’t try to exhume the great EJ and have a go at him.
The sooner this bloke retires the better. It will be good riddance

I read the top 50. Who cares really who is placed on the list and where they are. Its some old man who isn’t that bright and incredible biased.

Cooney isn’t that good of a player. He shouldn’t have polled so well in the last part of this year.

Chris Judd should be in front of Fevola. I agree. But Fevola can kick…and very well. Some of his goals are magic.

My rants:

Kicking 100 goals in a season does not make you an automatic number 1 player. If it were true it you should be arguing that Fevola should be second behind Franklin.

Thanks BN. Some early thoughts:

1. Gary Ablett - I tend to agree. Whilst Franklin can be much more damaging, his occassional inaccuracies reduce his value somewhat.
2. Lance Franklin - Will be interesting next year whether he plays as prominant a role, or whether he moves to a more periphary forward/midfield role.
3. Joel Corey - Perhaps a little high, but glad he’s ahead of JB.
4. Luke Hodge - Quite simply the best “footballer” around. If only the OP never happened. #1 defender listed.
5. Matthew Scarlett - Didn’t actually think his season was all that great - but still deserves to be the #2 (after Hodge).
6. Brent Harvey - Personally would have swapped with Corey.
7. J.Bartel - Too high for mine. Would have him below a few others.
8. B.Fevola - Despite only kicking a couple less goals than Franklin needs to be higher.
9. A.Cooney - Lower.
10. Chris Judd - Higher
11. Sam Fisher - Good to see. Been VERY good this year
12. Matthew Richardson - Deserved to be at 12.
13. Matthew Pavlich - Too low. He may have missed quite a few “game-deciding” shots, but he basically IS Fremantle. Should be top-10, maybe top-5.
14. Nick Riewoldt - Has he finally learnt to kick? Always talented, now Gehrig’s gone he might play closer to goal, where he can be much more dangerous than up on the wing.
15. Dean Cox - Has battled injuries all year, but performed well. About right.
16. Joel Selwood - Possibly a little low, I’d be happy to swap him and Bartel over.
17. Simon Black - A little low, 12-15 for mine.
18. Jonathan Brown - To be honest haven’t heard/seen much from Brown this year. Is he being rated on name/media rather than performance? It wouldn’t be the first time. About 25-30 for mine.
19. Sam Mitchell - About right. Can still have his influence reduced by great taggers.
20. Tom Harley - A little high, but not as bad a decision as I thought a week ago. 30ish.
21. Daniel Cross - If he could kick would be top 10. Rate him highly, but about right for his season.
22. Cameron Ling - Possibly a couple of spots higher, but happy enough.
23. Max Hudghton - About right. Best shut-down defender - knows how to use the umpires to his advantage.
24. Nathan Bock - A little low, would have him in the teens. Maybe swap with Brown.
25. Andrew Mackie - Probably a little high. 30ish
26. Chance Bateman - Really should be higher - much more important to Hawthorn than made out. I’d actually swap him and Sammy around.
27. Daniel Bradshaw - A little low, maybe in the high teens.
28. Drew Petrie
29. Corey Enright
30. Chad Cornes
31. Paul Medhurst - A little low. Had a great season.
32. Steve Johnson
33. Jarrod Roughead - Too low. Should be mid-20s, maybe even higher.
34. Dale Morris
35. Ryan O’Keefe
36. Leon Davis
37. Daniel Motlop
38. Aaron Sandilands - With his physical size, he really should be top 20 every year.
39. Brad Sewell
40. Robert Murphy
41. Brett Kirk - Too low for mine. 30ish.
42. Brett Deledio - Had a great year, probably pushing about ten spots too low.
43. Jarrad McVeigh
44. Dane Swan
45. Lenny Hayes
46. Luke Power
47. Scott Thompson
48. Michael Osbourne - A good selection - will be in top 5 at Hawthorn B&F and has had a very good year.
49. David Hille - Too low, should be around the 35 mark.
50. Matthew Boyd