prevayler-coders Mailing List for Prevayler (Page 4)
Brought to you by:
jsampson,
klauswuestefeld
Menu
▾
▴
prevayler-coders — People interested in the inner workings of the Prevayler project.
You can subscribe to this list here.
| 2004 |
Jan
|
Feb
|
Mar
|
Apr
(13) |
May
(83) |
Jun
(8) |
Jul
(18) |
Aug
(33) |
Sep
(132) |
Oct
(16) |
Nov
(32) |
Dec
(32) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2005 |
Jan
(35) |
Feb
(21) |
Mar
(7) |
Apr
(11) |
May
(18) |
Jun
|
Jul
(21) |
Aug
(14) |
Sep
(7) |
Oct
|
Nov
(16) |
Dec
|
| 2006 |
Jan
(15) |
Feb
|
Mar
|
Apr
|
May
|
Jun
(57) |
Jul
|
Aug
(15) |
Sep
(3) |
Oct
(1) |
Nov
|
Dec
|
| 2007 |
Jan
|
Feb
(1) |
Mar
|
Apr
|
May
|
Jun
(9) |
Jul
(7) |
Aug
|
Sep
(2) |
Oct
(1) |
Nov
|
Dec
(1) |
| 2008 |
Jan
|
Feb
(4) |
Mar
(10) |
Apr
(20) |
May
|
Jun
(5) |
Jul
(34) |
Aug
(2) |
Sep
|
Oct
(2) |
Nov
(2) |
Dec
|
| 2009 |
Jan
|
Feb
|
Mar
(5) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Klaus W. <kla...@gm...> - 2008-03-24 20:06:27
|
Git sounds nice. Isnt it better we move back from the API with generics, though? See you, Klaus. On Sun, Mar 23, 2008 at 3:21 AM, Justin T. Sampson <ju...@kr...> wrote: > Sorry. :) Git is yet another source control tool. Linus Torvalds wrote > it two years ago for Linux kernel development, when they had to move > away from the tool they'd been using (BitKeeper) for licensing > reasons. He hates CVS, and since Subversion aims to be "CVS done > right" he hates Subversion even more. Git's primary benefit is being > "distributed", meaning it doesn't require a central repository: Every > developer gets their own local writable copy of the entire repository, > with history, and Git handles merging changes better, reportedly, than > any other system. It seems to be gaining steam, and it definitely > feels very nice. I've just started playing with it for very small > things. > > http://git.or.cz/ > > Cheers, > Justin > > > > > On Sat, Mar 22, 2008 at 10:05 PM, Jacob Kjome <ho...@vi...> wrote: > > > Cool. Glad that's finally going to be done. BTW, what is "Git"? I supposed I > > could look it up on any search engine, but maybe you could provide a quick synopsis? > > > > Jake > > > > > > > > Justin T. Sampson wrote: > > > Hi Jake, > > > > > > I went ahead and requested the conversion: > > > > > > http://jira.codehaus.org/browse/HAUS-1651 > > > > > > I definitely want to get at the history myself. > > > > > > Was Prevayler 1 development in that same CVS repository, or hosted elsewhere? > > > > > > I've been getting interested in Git lately, and I'm tempted to import > > > the Codehaus CVS history into Git and then apply the Prevayler.org > > > Subversion work in a Git branch to make it easier to merge in bits and > > > pieces. Git would also make it easier to replicate the repository for > > > redundancy against these kinds of hosting maneuverings getting in the > > > way of commit access. We'll see... > > > > > > Cheers, > > > Justin > > > > > > > > > On Sun, Feb 17, 2008 at 10:32 AM, Jacob Kjome <ho...@vi...> wrote: > > > > > >> Since it's been decided that future development of Prevayler will happen off the > > >> 2.3 codbase instead of the experimental 3.0 codebase hosted by Klaus, we need to > > >> request that the 2.3 codebase be converted from its current CVS repository to SVN. > > >> This needs to be a request by the Despot. I guess that would be Klaus, right? > > >> > > >> Here's a message from Ben Walding on what needs to be done (and why). I would > > >> think the request would come in the form of a new Codehaus "chore" [1]. Until > > >> this is complete, any talk of Prevayler enhancements will be held up because of > > >> the frozen status of the Prevayler source repository. > > >> > > >> <quote name="Ben Walding"> > > >> We turned off CVS write access about 6 months ago during a migration of SCM > > >> services to a new host. It was just getting too annoying to provide SSH access on > > >> the required hosts. > > >> > > >> Despots can request conversion of their source repositories to SVN; this will let > > >> you resume committing to the repository. > > >> > > >> If the despots are MIA; we will convert the repository to SVN anyway; as we have > > >> no intention of reenabling CVS write access. > > >> </quote> > > >> > > >> [1] http://jira.codehaus.org/browse/HAUS > > >> > > >> > > >> Jake > > > > > > > > > > ------------------------------------------------------------------------- > > > This SF.net email is sponsored by: Microsoft > > > Defy all challenges. Microsoft(R) Visual Studio 2008. > > > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > > > _______________________________________________ > > > Prevayler-coders mailing list > > > Pre...@li... > > > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > > > > > > > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by: Microsoft > > Defy all challenges. Microsoft(R) Visual Studio 2008. > > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > > _______________________________________________ > > Prevayler-coders mailing list > > Pre...@li... > > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > |
|
From: Justin T. S. <ju...@kr...> - 2008-03-24 08:57:44
|
Prevayler's old Codehaus CVS repository has been converted to Subversion. Instructions here: http://xircles.codehaus.org/projects/prevayler/repo Cheers, Justin |
|
From: Justin T. S. <ju...@kr...> - 2008-03-23 06:21:05
|
Sorry. :) Git is yet another source control tool. Linus Torvalds wrote it two years ago for Linux kernel development, when they had to move away from the tool they'd been using (BitKeeper) for licensing reasons. He hates CVS, and since Subversion aims to be "CVS done right" he hates Subversion even more. Git's primary benefit is being "distributed", meaning it doesn't require a central repository: Every developer gets their own local writable copy of the entire repository, with history, and Git handles merging changes better, reportedly, than any other system. It seems to be gaining steam, and it definitely feels very nice. I've just started playing with it for very small things. http://git.or.cz/ Cheers, Justin On Sat, Mar 22, 2008 at 10:05 PM, Jacob Kjome <ho...@vi...> wrote: > Cool. Glad that's finally going to be done. BTW, what is "Git"? I supposed I > could look it up on any search engine, but maybe you could provide a quick synopsis? > > Jake > > > > Justin T. Sampson wrote: > > Hi Jake, > > > > I went ahead and requested the conversion: > > > > http://jira.codehaus.org/browse/HAUS-1651 > > > > I definitely want to get at the history myself. > > > > Was Prevayler 1 development in that same CVS repository, or hosted elsewhere? > > > > I've been getting interested in Git lately, and I'm tempted to import > > the Codehaus CVS history into Git and then apply the Prevayler.org > > Subversion work in a Git branch to make it easier to merge in bits and > > pieces. Git would also make it easier to replicate the repository for > > redundancy against these kinds of hosting maneuverings getting in the > > way of commit access. We'll see... > > > > Cheers, > > Justin > > > > > > On Sun, Feb 17, 2008 at 10:32 AM, Jacob Kjome <ho...@vi...> wrote: > > > >> Since it's been decided that future development of Prevayler will happen off the > >> 2.3 codbase instead of the experimental 3.0 codebase hosted by Klaus, we need to > >> request that the 2.3 codebase be converted from its current CVS repository to SVN. > >> This needs to be a request by the Despot. I guess that would be Klaus, right? > >> > >> Here's a message from Ben Walding on what needs to be done (and why). I would > >> think the request would come in the form of a new Codehaus "chore" [1]. Until > >> this is complete, any talk of Prevayler enhancements will be held up because of > >> the frozen status of the Prevayler source repository. > >> > >> <quote name="Ben Walding"> > >> We turned off CVS write access about 6 months ago during a migration of SCM > >> services to a new host. It was just getting too annoying to provide SSH access on > >> the required hosts. > >> > >> Despots can request conversion of their source repositories to SVN; this will let > >> you resume committing to the repository. > >> > >> If the despots are MIA; we will convert the repository to SVN anyway; as we have > >> no intention of reenabling CVS write access. > >> </quote> > >> > >> [1] http://jira.codehaus.org/browse/HAUS > >> > >> > >> Jake > > > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by: Microsoft > > Defy all challenges. Microsoft(R) Visual Studio 2008. > > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > > _______________________________________________ > > Prevayler-coders mailing list > > Pre...@li... > > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > |
|
From: Jacob K. <ho...@vi...> - 2008-03-23 04:04:15
|
Cool. Glad that's finally going to be done. BTW, what is "Git"? I supposed I could look it up on any search engine, but maybe you could provide a quick synopsis? Jake Justin T. Sampson wrote: > Hi Jake, > > I went ahead and requested the conversion: > > http://jira.codehaus.org/browse/HAUS-1651 > > I definitely want to get at the history myself. > > Was Prevayler 1 development in that same CVS repository, or hosted elsewhere? > > I've been getting interested in Git lately, and I'm tempted to import > the Codehaus CVS history into Git and then apply the Prevayler.org > Subversion work in a Git branch to make it easier to merge in bits and > pieces. Git would also make it easier to replicate the repository for > redundancy against these kinds of hosting maneuverings getting in the > way of commit access. We'll see... > > Cheers, > Justin > > > On Sun, Feb 17, 2008 at 10:32 AM, Jacob Kjome <ho...@vi...> wrote: > >> Since it's been decided that future development of Prevayler will happen off the >> 2.3 codbase instead of the experimental 3.0 codebase hosted by Klaus, we need to >> request that the 2.3 codebase be converted from its current CVS repository to SVN. >> This needs to be a request by the Despot. I guess that would be Klaus, right? >> >> Here's a message from Ben Walding on what needs to be done (and why). I would >> think the request would come in the form of a new Codehaus "chore" [1]. Until >> this is complete, any talk of Prevayler enhancements will be held up because of >> the frozen status of the Prevayler source repository. >> >> <quote name="Ben Walding"> >> We turned off CVS write access about 6 months ago during a migration of SCM >> services to a new host. It was just getting too annoying to provide SSH access on >> the required hosts. >> >> Despots can request conversion of their source repositories to SVN; this will let >> you resume committing to the repository. >> >> If the despots are MIA; we will convert the repository to SVN anyway; as we have >> no intention of reenabling CVS write access. >> </quote> >> >> [1] http://jira.codehaus.org/browse/HAUS >> >> >> Jake > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > |
|
From: Justin T. S. <ju...@kr...> - 2008-03-22 04:59:56
|
Hi Jake, I went ahead and requested the conversion: http://jira.codehaus.org/browse/HAUS-1651 I definitely want to get at the history myself. Was Prevayler 1 development in that same CVS repository, or hosted elsewhere? I've been getting interested in Git lately, and I'm tempted to import the Codehaus CVS history into Git and then apply the Prevayler.org Subversion work in a Git branch to make it easier to merge in bits and pieces. Git would also make it easier to replicate the repository for redundancy against these kinds of hosting maneuverings getting in the way of commit access. We'll see... Cheers, Justin On Sun, Feb 17, 2008 at 10:32 AM, Jacob Kjome <ho...@vi...> wrote: > Since it's been decided that future development of Prevayler will happen off the > 2.3 codbase instead of the experimental 3.0 codebase hosted by Klaus, we need to > request that the 2.3 codebase be converted from its current CVS repository to SVN. > This needs to be a request by the Despot. I guess that would be Klaus, right? > > Here's a message from Ben Walding on what needs to be done (and why). I would > think the request would come in the form of a new Codehaus "chore" [1]. Until > this is complete, any talk of Prevayler enhancements will be held up because of > the frozen status of the Prevayler source repository. > > <quote name="Ben Walding"> > We turned off CVS write access about 6 months ago during a migration of SCM > services to a new host. It was just getting too annoying to provide SSH access on > the required hosts. > > Despots can request conversion of their source repositories to SVN; this will let > you resume committing to the repository. > > If the despots are MIA; we will convert the repository to SVN anyway; as we have > no intention of reenabling CVS write access. > </quote> > > [1] http://jira.codehaus.org/browse/HAUS > > > Jake |
|
From: Justin T. S. <ju...@kr...> - 2008-02-18 09:33:05
|
Hi Jake! I'll respond to what I can... On Feb 17, 2008 10:02 PM, Jacob Kjome <ho...@vi...> wrote: > It's fine if source control for Prevayler2 is moved to prevayler.org. I'm entirely agnostic about it myself. > However, I > beg to differ that the link you provided represents the prevayler2 codebase. True enough. I think the initial revision in that Subversion repository should match the head revision in the Codehaus CVS repository, though. > It > looks to be significantly modified (are the changes detailed anywhere, BTW?). The head revision is quite a bit modified, though in a sense it's mostly cosmetic -- the internal mechanisms are all pretty much the same as Prevayler 2.3. I was experimenting with different ways of leveraging Java 5 generics and annotations. The Subversion commit comments should give some sense of the kinds of changes I was making. I put a smallish application in production based on this latest code; I'm confident in its quality but in the end I came to the conclusion that Java 5 generics didn't make my code any cleaner (in fact, they were quite a bit of a pain), so none of these changes is much worth keeping. I may have made some tweaks around things like read-write locking that might be interesting to some users (I know I had it in mind because Java 5's new concurrency util package offers a read-write lock implementation), but I don't remember if I actually got to that kind of thing. > Additionally, there is no history for the older code. That is, the previous tags > and branches have been discarded. Right. Well, not so much "discarded", as "not imported". > If this is the repository of record, it's not a > very well kept one. Even if you want to start fresh on the next Prevayler2 > release, the older tagged releases and branches should still be imported and > stored for posterity (the Prevayler1 source ought to be made available as well). Sounds good to me. For the record, I don't have the time or Subversion skills to do it myself. If Codehaus is offering to do it over there, that sounds pretty cool. > Also, what's with the mime types for stuff like GIF images or Java source? Trying > to view them in a web browser causes a download dialog to be presented. I think > your web server needs to be updated with a proper list of mime types so that > browsers know what to do with them. > > BTW, do I still have commit rights? I used to. These are purely server config questions that I'll leave to Klaus. Cheers, Justin |
|
From: Jacob K. <ho...@vi...> - 2008-02-18 06:01:53
|
Hi Klaus, It's fine if source control for Prevayler2 is moved to prevayler.org. However, I beg to differ that the link you provided represents the prevayler2 codebase. It looks to be significantly modified (are the changes detailed anywhere, BTW?). Additionally, there is no history for the older code. That is, the previous tags and branches have been discarded. If this is the repository of record, it's not a very well kept one. Even if you want to start fresh on the next Prevayler2 release, the older tagged releases and branches should still be imported and stored for posterity (the Prevayler1 source ought to be made available as well). Also, what's with the mime types for stuff like GIF images or Java source? Trying to view them in a web browser causes a download dialog to be presented. I think your web server needs to be updated with a proper list of mime types so that browsers know what to do with them. BTW, do I still have commit rights? I used to. Jake Klaus Wuestefeld wrote: > Hi, > > Prevayler 2 SVN (official head) is here: > http://prevayler.org/svn/prevayler/trunk/java/prevayler2/ > > My Prevayler 3 musings are here: > http://prevayler.org/svn/prevayler/trunk/java/prevayler3/ > > The codehaus CVS can be gc'd. > > Thanks. > > > > On Feb 17, 2008 3:32 PM, Jacob Kjome <ho...@vi...> wrote: >> Since it's been decided that future development of Prevayler will happen off the >> 2.3 codbase instead of the experimental 3.0 codebase hosted by Klaus, we need to >> request that the 2.3 codebase be converted from its current CVS repository to SVN. >> This needs to be a request by the Despot. I guess that would be Klaus, right? >> >> Here's a message from Ben Walding on what needs to be done (and why). I would >> think the request would come in the form of a new Codehaus "chore" [1]. Until >> this is complete, any talk of Prevayler enhancements will be held up because of >> the frozen status of the Prevayler source repository. >> >> <quote name="Ben Walding"> >> We turned off CVS write access about 6 months ago during a migration of SCM >> services to a new host. It was just getting too annoying to provide SSH access on >> the required hosts. >> >> Despots can request conversion of their source repositories to SVN; this will let >> you resume committing to the repository. >> >> If the despots are MIA; we will convert the repository to SVN anyway; as we have >> no intention of reenabling CVS write access. >> </quote> >> >> [1] http://jira.codehaus.org/browse/HAUS >> >> >> Jake >> >> ------------------------------------------------------------------------- >> This SF.net email is sponsored by: Microsoft >> Defy all challenges. Microsoft(R) Visual Studio 2008. >> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ >> _______________________________________________ >> Prevayler-coders mailing list >> Pre...@li... >> https://lists.sourceforge.net/lists/listinfo/prevayler-coders >> > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > |
|
From: Klaus W. <kla...@gm...> - 2008-02-18 03:50:33
|
Hi, Prevayler 2 SVN (official head) is here: http://prevayler.org/svn/prevayler/trunk/java/prevayler2/ My Prevayler 3 musings are here: http://prevayler.org/svn/prevayler/trunk/java/prevayler3/ The codehaus CVS can be gc'd. Thanks. On Feb 17, 2008 3:32 PM, Jacob Kjome <ho...@vi...> wrote: > Since it's been decided that future development of Prevayler will happen off the > 2.3 codbase instead of the experimental 3.0 codebase hosted by Klaus, we need to > request that the 2.3 codebase be converted from its current CVS repository to SVN. > This needs to be a request by the Despot. I guess that would be Klaus, right? > > Here's a message from Ben Walding on what needs to be done (and why). I would > think the request would come in the form of a new Codehaus "chore" [1]. Until > this is complete, any talk of Prevayler enhancements will be held up because of > the frozen status of the Prevayler source repository. > > <quote name="Ben Walding"> > We turned off CVS write access about 6 months ago during a migration of SCM > services to a new host. It was just getting too annoying to provide SSH access on > the required hosts. > > Despots can request conversion of their source repositories to SVN; this will let > you resume committing to the repository. > > If the despots are MIA; we will convert the repository to SVN anyway; as we have > no intention of reenabling CVS write access. > </quote> > > [1] http://jira.codehaus.org/browse/HAUS > > > Jake > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > |
|
From: Jacob K. <ho...@vi...> - 2008-02-17 17:32:22
|
Since it's been decided that future development of Prevayler will happen off the
2.3 codbase instead of the experimental 3.0 codebase hosted by Klaus, we need to
request that the 2.3 codebase be converted from its current CVS repository to SVN.
This needs to be a request by the Despot. I guess that would be Klaus, right?
Here's a message from Ben Walding on what needs to be done (and why). I would
think the request would come in the form of a new Codehaus "chore" [1]. Until
this is complete, any talk of Prevayler enhancements will be held up because of
the frozen status of the Prevayler source repository.
<quote name="Ben Walding">
We turned off CVS write access about 6 months ago during a migration of SCM
services to a new host. It was just getting too annoying to provide SSH access on
the required hosts.
Despots can request conversion of their source repositories to SVN; this will let
you resume committing to the repository.
If the despots are MIA; we will convert the repository to SVN anyway; as we have
no intention of reenabling CVS write access.
</quote>
[1] http://jira.codehaus.org/browse/HAUS
Jake
|
|
From: Klaus W. <kla...@gm...> - 2007-12-23 17:24:39
|
Hi, Our svn servers and website will be down for a few days for maintenance purposes. See you, Klaus. |
|
From: Brian S. <br...@sl...> - 2007-10-21 06:29:36
|
"In the late 1970's a large machine had somewhere around a megabyte of main memory. Today, several Gbytes are common and large machines are approaching 100 Gbytes. In a few years a terabyte of main memory will not be unusual. Imagine a shared nothing grid system of 20 nodes, each with 32 Gbytes of main memory now, (soon to be 100 Gbytes), and costing less than $50,000. As such, any database less than a terabyte in size, is capable of main memory deployment now or in the near future." The End of an Architectural Era (It's Time for a Complete Rewrite) http://lambda-the-ultimate.org/node/2500 |
|
From: Klaus W. <kla...@gm...> - 2007-09-25 14:08:52
|
VERY, VERY NICE. :) A put a comment at the end. See you, Klaus. On 9/25/07, Justin T. Sampson <ju...@kr...> wrote: > Howdy, > > I finally got around to working a bit on the Prevayler wiki. Actually, > this week is the last week of my company's fiscal year, and I had 2 > more hours of volunteer work to do, so I had a bit of impetus. My > bonus is dependent on it! And then I'll have more time starting next > week (10 days per year). > > http://www.prevayler.org/ > > It might look small (just some new text and a diagram on the front > page), but there's a bit of organizing work that went into it. I don't > really want to restore the text of the old wiki; but I've categorized > it all to make sure I don't miss anything as I'm building out the new > wiki. I'll take bits of text as appropriate. For the front page, I > really wanted to dive right in with a description of how Prevayler > works, since many people just don't get the basics. I think it's cool > that the entire architecture can be described pretty well in a single > paragraph, and that's one of the best ways to present Prevayler to the > world. :) And some of the most common questions are around "what rules > do I have to follow", so I put that right there on the homepage as > well. I also wanted to make the reasons for using Prevayler more > personal -- rather than making an a priori claim about what "you" will > be able to do, which I think puts some people off, I wanted to make it > clear that these are real benefits that we have really gotten > ourselves. > > I imagine each of the parts of the diagram in the first section will > have its own page with further details, as will each of the "why's" > and each of the "rules". There's lots to pull from the old wiki, such > as more detail on the various things that aren't deterministic > (hashCode, etc.). In addition, there are some broad categories with a > lot of content in the old wiki, such as philosophical background and > implications. These are the rough categories I've identified: > > PRINCIPLES (philosophy, background, implications) > RULES (how to use Prevayler) > REAL FEATURES (already implemented) > DREAM FEATURES (future, maybe) > DEMOS (Bank etc.) > USAGES (real-world production apps) > SCALABILITY TEST > PATTERNS ("how do I do *this*?") > "IS NOT" ("Prevayler is not a ...") > REFERENCES (articles, etc.) > PORTS (Bamboo, Florypa, etc.) > > As I categorized the old content I also thought of some things that > weren't ever mentioned on the old wiki but probably should be -- some > rules, some patterns, that kind of thing. > > By the way, there are some little quirks in the new wiki software. > It's adding spaces around all my apostrophes and double-quotes, so > "It's" becomes "It ' s", for example. Also, I couldn't find a way to > upload an image, so I just put the diagram up on my own server and > linked to it there. Overall it looks really nice, though! > > Cheers, > Justin > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2005. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > |
|
From: Justin T. S. <ju...@kr...> - 2007-09-25 03:51:23
|
Howdy, I finally got around to working a bit on the Prevayler wiki. Actually, this week is the last week of my company's fiscal year, and I had 2 more hours of volunteer work to do, so I had a bit of impetus. My bonus is dependent on it! And then I'll have more time starting next week (10 days per year). http://www.prevayler.org/ It might look small (just some new text and a diagram on the front page), but there's a bit of organizing work that went into it. I don't really want to restore the text of the old wiki; but I've categorized it all to make sure I don't miss anything as I'm building out the new wiki. I'll take bits of text as appropriate. For the front page, I really wanted to dive right in with a description of how Prevayler works, since many people just don't get the basics. I think it's cool that the entire architecture can be described pretty well in a single paragraph, and that's one of the best ways to present Prevayler to the world. :) And some of the most common questions are around "what rules do I have to follow", so I put that right there on the homepage as well. I also wanted to make the reasons for using Prevayler more personal -- rather than making an a priori claim about what "you" will be able to do, which I think puts some people off, I wanted to make it clear that these are real benefits that we have really gotten ourselves. I imagine each of the parts of the diagram in the first section will have its own page with further details, as will each of the "why's" and each of the "rules". There's lots to pull from the old wiki, such as more detail on the various things that aren't deterministic (hashCode, etc.). In addition, there are some broad categories with a lot of content in the old wiki, such as philosophical background and implications. These are the rough categories I've identified: PRINCIPLES (philosophy, background, implications) RULES (how to use Prevayler) REAL FEATURES (already implemented) DREAM FEATURES (future, maybe) DEMOS (Bank etc.) USAGES (real-world production apps) SCALABILITY TEST PATTERNS ("how do I do *this*?") "IS NOT" ("Prevayler is not a ...") REFERENCES (articles, etc.) PORTS (Bamboo, Florypa, etc.) As I categorized the old content I also thought of some things that weren't ever mentioned on the old wiki but probably should be -- some rules, some patterns, that kind of thing. By the way, there are some little quirks in the new wiki software. It's adding spaces around all my apostrophes and double-quotes, so "It's" becomes "It ' s", for example. Also, I couldn't find a way to upload an image, so I just put the diagram up on my own server and linked to it there. Overall it looks really nice, though! Cheers, Justin |
|
From: Klaus W. <kla...@gm...> - 2007-07-23 17:09:23
|
> It seems like there is a possible case of single threaded programs > that use frequent transactions and tolerate loss of the last > transactions on rollback, which would be best served by support for > asynchronous writing. I believe so too. As for Maven packaging, if you do it, I will talk to the iBiblio guys. See you, Klaus. On 7/23/07, Hans H=E4ggstr=F6m <han...@gm...> wrote: > Hi Justin, thanks for responding! > > On 7/22/07, Justin T. Sampson <ju...@kr...> wrote: > <snip> > > > > The biggest impact of this batching technique comes when you have a > > relatively large number of client threads running. With a single > > client thread of course you wouldn't notice it. > > > > How many client > > threads do you have running? > > I'm using a single thread for doing the game world simulation. Later > there would probably be more threads to handle the connected game > clients, when I get there I may have to revise the threading model for > the game. > > I'd prefer to use just one thread for doing the game state updates > though, to avoid complicating the code. > > Even if I used a multithreaded implementation, asynchronous disk > writing could still improve the performance, unless the batched > updates became very large. > > > > How many transactions do they each > > generate per second? > > I'm sending about 10 world update clock ticks per second now as transacti= ons. > > I might be able to later embed > number_of_world_ticks_since_last_transaction into transactions from > the game clients, thus eliminating the need for separate clock update > transactions. Still, with 50 - 100 players or so, the number of > incoming player action transactions would reach similar frequencies > (10 to 100 updates / second). > > > > What kind of throughput (total transactions per > > second) were you getting before making your change? What kind of > > throughput are you getting after making your change? > > The difference was quite noticeable, the speedup was probably in the > order of 100 - 1000%. > I'm moving my build system to maven at the moment, so I don't have a > compiling test program at hand. I can post comparisons later when I > get things running again. > > (Speaking of that, the Prevayler version in the Maven2 repository > seems to be out of date, I can't see Prevayler 2.3 there: > http://mirrors.ibiblio.org/pub/mirrors/maven2/prevayler/prevayler/ > The uploading guide is here: > http://maven.apache.org/guides/mini/guide-central-repository-upload.html > (if needed I can try to help with the packaging, but I don't have > access to ibiblio)) > > > In addition, I'm a bit worried about the abuse on the disk drive - > with synchronous writing, it was making continuous noise, as opposed > to the occasional write burst with asynchronous writing. Seems like > synchronous writing with frequent transactions could burn out a disk > much faster on a server machine, but then I'm no expert on hard > drives. > > > > Is your total > > system memory usage any different, due to keeping more transactions in > > the disk buffer before writing? > > I doubt that would have any great impact on the system memory usage, > my transactions are not big (maybe 50 - 200 bytes typically), and with > 100 of them in a second, it still would stay easily under 1 MB unless > the operating system waited more than a minute or so between writing > things to disk. > > > > (We actually had an option like the one you've implemented in an > > earlier version of Prevayler, but dropped it in favor of batched > > writes, which provide much of the performance benefits without the > > safety drawbacks.) > > It seems like there is a possible case of single threaded programs > that use frequent transactions and tolerate loss of the last > transactions on rollback, which would be best served by support for > asynchronous writing. > > > -- Hans > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > |
|
From: <han...@gm...> - 2007-07-23 09:53:12
|
Hi Justin, thanks for responding! On 7/22/07, Justin T. Sampson <ju...@kr...> wrote: <snip> > > The biggest impact of this batching technique comes when you have a > relatively large number of client threads running. With a single > client thread of course you wouldn't notice it. > > How many client > threads do you have running? I'm using a single thread for doing the game world simulation. Later there would probably be more threads to handle the connected game clients, when I get there I may have to revise the threading model for the game. I'd prefer to use just one thread for doing the game state updates though, to avoid complicating the code. Even if I used a multithreaded implementation, asynchronous disk writing could still improve the performance, unless the batched updates became very large. > How many transactions do they each > generate per second? I'm sending about 10 world update clock ticks per second now as transactions. I might be able to later embed number_of_world_ticks_since_last_transaction into transactions from the game clients, thus eliminating the need for separate clock update transactions. Still, with 50 - 100 players or so, the number of incoming player action transactions would reach similar frequencies (10 to 100 updates / second). > What kind of throughput (total transactions per > second) were you getting before making your change? What kind of > throughput are you getting after making your change? The difference was quite noticeable, the speedup was probably in the order of 100 - 1000%. I'm moving my build system to maven at the moment, so I don't have a compiling test program at hand. I can post comparisons later when I get things running again. (Speaking of that, the Prevayler version in the Maven2 repository seems to be out of date, I can't see Prevayler 2.3 there: http://mirrors.ibiblio.org/pub/mirrors/maven2/prevayler/prevayler/ The uploading guide is here: http://maven.apache.org/guides/mini/guide-central-repository-upload.html (if needed I can try to help with the packaging, but I don't have access to ibiblio)) In addition, I'm a bit worried about the abuse on the disk drive - with synchronous writing, it was making continuous noise, as opposed to the occasional write burst with asynchronous writing. Seems like synchronous writing with frequent transactions could burn out a disk much faster on a server machine, but then I'm no expert on hard drives. > Is your total > system memory usage any different, due to keeping more transactions in > the disk buffer before writing? I doubt that would have any great impact on the system memory usage, my transactions are not big (maybe 50 - 200 bytes typically), and with 100 of them in a second, it still would stay easily under 1 MB unless the operating system waited more than a minute or so between writing things to disk. > (We actually had an option like the one you've implemented in an > earlier version of Prevayler, but dropped it in favor of batched > writes, which provide much of the performance benefits without the > safety drawbacks.) It seems like there is a possible case of single threaded programs that use frequent transactions and tolerate loss of the last transactions on rollback, which would be best served by support for asynchronous writing. -- Hans |
|
From: Brian S. <br...@sl...> - 2007-07-22 19:45:23
|
>From the original message, it sounds like there is a clock tick for the entire game world, so it's effectively single-threaded. I suppose a game simulation might be done that's not based on global clock ticks, but that's a fundamental change in architecture, and probably much more complicated. - Brian On 7/22/07, Justin T. Sampson <ju...@kr...> wrote: > Hi Hans, > > Thanks for contributing! > > Are you aware of how Prevayler currently batches transactions to be > written to the journal? For correctness, each individual client thread > does indeed wait for its transaction to be written before continuing, > but if several client threads are submitting transactions > concurrently, they will be batched into a single disk write. The > current code handles this very dynamically, simply batching > transactions until the previous write has completed, to maximize > throughput to the disk. > > The biggest impact of this batching technique comes when you have a > relatively large number of client threads running. With a single > client thread of course you wouldn't notice it. How many client > threads do you have running? How many transactions do they each > generate per second? What kind of throughput (total transactions per > second) were you getting before making your change? What kind of > throughput are you getting after making your change? Is your total > system memory usage any different, due to keeping more transactions in > the disk buffer before writing? > > (We actually had an option like the one you've implemented in an > earlier version of Prevayler, but dropped it in favor of batched > writes, which provide much of the performance benefits without the > safety drawbacks.) > > Cheers, > Justin > > > On 7/21/07, Hans H=E4ggstr=F6m <han...@gm...> wrote: > > > Hello, > > > > Attached is a patch to add support for asynchronous disk writing to Pre= vayler. > > > > > > Motivation: > > > > Normally Prevayler writes the transactions synchronously to disk. > > That means that the thread won't return from the I/O operation before > > the transaction has been physically written to the disk. This ensures > > reliability, in that once a transaction has been written, it can be > > rolled back. > > > > However, in some applications performance is more important than high > > reliability like this. If the number of transactions per second is > > significant, the synchronous writing will slow down the execution > > significantly (in addition to abusing the harddrive with a lot of > > frequent short writes). > > > > I'm using this for a persistent RPG game server that I'm working on, > > where transactions are incoming player actions and world simulation > > clock ticks. In this application, it's okay if the transactions from > > the last few minutes are lost in the event of a server crash, if the > > system can perform faster because of it. > > > > > > Changes: > > > > The patch enables the user to select between asynchronous and > > synchronous disk writing, using the > > PrevaylerFactory.configureSyncronousDiskWriting method (the default is > > still synchronous like before). Asynchronous disk writing uses Javas > > normal file flushing, which will write changes to the disk in batches > > typically every few seconds or so (depending on operating system > > settings). > > > > The patch modifies three classes: PrevaylerFactory, PersistentJournal, > > and DurableOutputStream. It adds some new methods and constructors, > > but doesn't otherwise change the API. The default behavior remains > > unchanged, so it should have no effect on existing applications. > > > > The patch is generated against the Prevayler 2.3 release sources, > > using unified diff format. > > > > > > -- Hans H=E4ggstr=F6m > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > |
|
From: Justin T. S. <ju...@kr...> - 2007-07-22 18:11:55
|
Hi Hans, Thanks for contributing! Are you aware of how Prevayler currently batches transactions to be written to the journal? For correctness, each individual client thread does indeed wait for its transaction to be written before continuing, but if several client threads are submitting transactions concurrently, they will be batched into a single disk write. The current code handles this very dynamically, simply batching transactions until the previous write has completed, to maximize throughput to the disk. The biggest impact of this batching technique comes when you have a relatively large number of client threads running. With a single client thread of course you wouldn't notice it. How many client threads do you have running? How many transactions do they each generate per second? What kind of throughput (total transactions per second) were you getting before making your change? What kind of throughput are you getting after making your change? Is your total system memory usage any different, due to keeping more transactions in the disk buffer before writing? (We actually had an option like the one you've implemented in an earlier version of Prevayler, but dropped it in favor of batched writes, which provide much of the performance benefits without the safety drawbacks.) Cheers, Justin On 7/21/07, Hans H=E4ggstr=F6m <han...@gm...> wrote: > Hello, > > Attached is a patch to add support for asynchronous disk writing to Preva= yler. > > > Motivation: > > Normally Prevayler writes the transactions synchronously to disk. > That means that the thread won't return from the I/O operation before > the transaction has been physically written to the disk. This ensures > reliability, in that once a transaction has been written, it can be > rolled back. > > However, in some applications performance is more important than high > reliability like this. If the number of transactions per second is > significant, the synchronous writing will slow down the execution > significantly (in addition to abusing the harddrive with a lot of > frequent short writes). > > I'm using this for a persistent RPG game server that I'm working on, > where transactions are incoming player actions and world simulation > clock ticks. In this application, it's okay if the transactions from > the last few minutes are lost in the event of a server crash, if the > system can perform faster because of it. > > > Changes: > > The patch enables the user to select between asynchronous and > synchronous disk writing, using the > PrevaylerFactory.configureSyncronousDiskWriting method (the default is > still synchronous like before). Asynchronous disk writing uses Javas > normal file flushing, which will write changes to the disk in batches > typically every few seconds or so (depending on operating system > settings). > > The patch modifies three classes: PrevaylerFactory, PersistentJournal, > and DurableOutputStream. It adds some new methods and constructors, > but doesn't otherwise change the API. The default behavior remains > unchanged, so it should have no effect on existing applications. > > The patch is generated against the Prevayler 2.3 release sources, > using unified diff format. > > > -- Hans H=E4ggstr=F6m |
|
From: <han...@gm...> - 2007-07-22 06:42:15
|
On 7/22/07, Brian Slesinsky <br...@sl...> wrote: > Thanks for the patch! I'm not sure anyone is working on Prevayler > right now, but the people who need it should be able to find it on the > mailing list. Ok. It seems like the download from the mailing list archives is not working at least for me, so I also uploaded the patch to my homepage at: http://www.iki.fi/zzorn/prevaylerasynchronouswritingpatch (currently redirecting to http://hans.haggstrom.googlepages.com/prevaylerasynchronouswritingpatch). Also, configureSyncronousDiskWriting should of course be named configureSynchronousDiskWriting (correct spelling), but I'm too lazy to update the patch with that at the moment. It's a one character change anyway. -- Hans |
|
From: Brian S. <br...@sl...> - 2007-07-21 23:39:59
|
Thanks for the patch! I'm not sure anyone is working on Prevayler right now, but the people who need it should be able to find it on the mailing list. - Brian |
|
From: <han...@gm...> - 2007-07-21 11:14:13
|
Hello, Attached is a patch to add support for asynchronous disk writing to Prevayl= er. Motivation: Normally Prevayler writes the transactions synchronously to disk. That means that the thread won't return from the I/O operation before the transaction has been physically written to the disk. This ensures reliability, in that once a transaction has been written, it can be rolled back. However, in some applications performance is more important than high reliability like this. If the number of transactions per second is significant, the synchronous writing will slow down the execution significantly (in addition to abusing the harddrive with a lot of frequent short writes). I'm using this for a persistent RPG game server that I'm working on, where transactions are incoming player actions and world simulation clock ticks. In this application, it's okay if the transactions from the last few minutes are lost in the event of a server crash, if the system can perform faster because of it. Changes: The patch enables the user to select between asynchronous and synchronous disk writing, using the PrevaylerFactory.configureSyncronousDiskWriting method (the default is still synchronous like before). Asynchronous disk writing uses Javas normal file flushing, which will write changes to the disk in batches typically every few seconds or so (depending on operating system settings). The patch modifies three classes: PrevaylerFactory, PersistentJournal, and DurableOutputStream. It adds some new methods and constructors, but doesn't otherwise change the API. The default behavior remains unchanged, so it should have no effect on existing applications. The patch is generated against the Prevayler 2.3 release sources, using unified diff format. -- Hans H=E4ggstr=F6m |
|
From: Klaus W. <kla...@gm...> - 2007-06-19 22:28:11
|
I didnt quite understand what you did. Could you describe it in a bit more detail? See you, Klaus. On 6/19/07, Travelmole Travelmole <tes...@ya...> wrote: > Hi, > > I made the tar file of all the prevayler data folder and moved at a new > machine. Deleted all the commanLog files and old snapshot files. And > restarted the web server. Fortunately I found all the data including the > new one. > > Will it be safe to restore with this new snapshots only? > > Pls sugess.. > > Thanks & Regards, > Om > > Klaus Wuestefeld <kla...@gm...> wrote: > No. All transactions after log 347 will not be in this snapshot. > > Brians guess is probably right. Something went wrong in log 347. If > you can send us the code and the logs we can take a look. > > See you, Klaus. > > > On 6/18/07, Travelmole Travelmole wrote: > > Hi, > > > > I have taken the snapshot, but 347 snapshot has been created. > > > > -rw-r--r-- 1 root root 197725 Jun 18 10:27 > > 000000000000000000347.snapshot > > > > Will this snapshot be having all the datas? > > > > Pls sugess me the next step to do. > > > > Thanks, > > Om > > > > > > Travelmole Travelmole wrote: > > Hi Klaus, > > > > Thanks for responce. > > > > Could you pls sugess me the step by step way to take a complete snapshot > and > > restore. > > > > Thanks & Regards, > > Om > > Klaus Wuestefeld wrote: > > Hi :) > > > > Simply take a snapshot and delete the log. That should do it. > > > > But can you send us the log and the code? I have never seen this case. > > > > Thanks, Klaus. > > > > > > > > On 6/16/07, Travelmole Travelmole wrote: > > > Hi everybody, > > > > > > I am using prevayler version 2.1 to store my data. The application has > > been > > > running for last 2 years. > > > > > > Since last month, new data created is being lost when we restart the web > > > server. > > > > > > In the catalina.out file, I am seeing multiple instances of this > exception > > - > > > > > > java.io.IOException: Attempt to create command log file that already > > > existed: /home/demo/prevayler/pro > > > jects/000000000000000000348.commandLog > > > at > > > > > > org.prevayler.implementation.NumberFileCreator.newLog(NumberFileCreator.java:28) > > > at > > > > > > org.prevayler.implementation.CommandOutputStream.logStream(CommandOutputStream.java:52) > > > at > > > > > > org.prevayler.implementation.CommandOutputStream.logStream(CommandOutputStream.java:58) > > > at > > > > > > org.prevayler.implementation.CommandOutputStream.writeCommand(CommandOutputStream.java:26) > > > at > > > > > > org.prevayler.implementation.SnapshotPrevayler.executeCommand(SnapshotPrevayler.java:71) > > > .... > > > .... > > > > > > I am using linux, Tomcat as web server. > > > > > > List of last few CommandLog file is as below ( note the sequence of file > > is > > > not in same sequence as dates !! ) - > > > > > > -rwxrw-rw- 1 root root 100554 Feb 8 03:08 > > > 000000000000000000341.commandLog > > > -rwxrw-rw- 1 root root 100490 Feb 8 03:08 > > > 000000000000000000342.commandLog > > > -rwxrw-rw- 1 root root 59599 Feb 8 03:08 > > > 000000000000000000343.commandLog > > > -rwxrw-rw- 1 root root 101121 May 25 08:35 > > > 000000000000000000344.commandLog > > > -rwxrw-rw- 1 root root 100992 May 25 08:35 > > > 000000000000000000345.commandLog > > > -rwxrw-rw- 1 root root 100690 May 25 08:35 > > > 000000000000000000346.commandLog > > > -rwxrw-rw- 1 root root 100901 Jun 8 05:26 > > > 000000000000000000347.commandLog > > > -rwxrw-rw- 1 root root 101121 May 25 09:06 > > > 000000000000000000348.commandLog > > > -rwxrw-rw- 1 root root 100992 May 25 09:06 > > > 000000000000000000349.commandLog > > > -rwxrw-rw- 1 root root 100690 May 25 09:06 > > > 000000000000000000350.commandLog > > > -rwxrw-rw- 1 root root 58317 May 25 09:06 > > > 000000000000000000351.commandLog > > > > > > How do I resolve this problem ? > > > > > > Thanks & Regards, > > > Om > > > > > > ________________________________ > > > Be a PS3 game guru. > > > Get your game face on with the latest PS3 news and previews at Yahoo! > > Games. > > > > > > > > > > > > ------------------------------------------------------------------------- > > > This SF.net email is sponsored by DB2 Express > > > Download DB2 Express C - the FREE version of DB2 express and take > > > control of your XML. No limits. Just data. Click to get it now. > > > http://sourceforge.net/powerbar/db2/ > > > _______________________________________________ > > > Prevayler-coders mailing list > > > Pre...@li... > > > > > > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > > > > > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by DB2 Express > > Download DB2 Express C - the FREE version of DB2 express and take > > control of your XML. No limits. Just data. Click to get it now. > > http://sourceforge.net/powerbar/db2/ > > _______________________________________________ > > Prevayler-coders mailing list > > Pre...@li... > > > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > ________________________________ > > Looking for a deal? Find great prices on flights and hotels with Yahoo! > > > FareChase.------------------------------------------------------------------------- > > This SF.net email is sponsored by DB2 Express > > Download DB2 Express C - the FREE version of DB2 express and take > > control of your XML. No limits. Just data. Click to get it now. > > > http://sourceforge.net/powerbar/db2/_______________________________________________ > > Prevayler-coders mailing list > > Pre...@li... > > > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > > > > > ________________________________ > > Yahoo! oneSearch: Finally, mobile search that gives answers, not web > links. > > > > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by DB2 Express > > Download DB2 Express C - the FREE version of DB2 express and take > > control of your XML. No limits. Just data. Click to get it now. > > http://sourceforge.net/powerbar/db2/ > > _______________________________________________ > > Prevayler-coders mailing list > > Pre...@li... > > > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > ________________________________ > Be a PS3 game guru. > Get your game face on with the latest PS3 news and previews at Yahoo! Games. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > |
|
From: Travelmole T. <tes...@ya...> - 2007-06-19 13:51:40
|
Hi, I made the tar file of all the prevayler data folder and moved at a new machine. Deleted all the commanLog files and old snapshot files. And restarted the web server. Fortunately I found all the data including the new one. Will it be safe to restore with this new snapshots only? Pls sugess.. Thanks & Regards, Om Klaus Wuestefeld <kla...@gm...> wrote: No. All transactions after log 347 will not be in this snapshot. Brians guess is probably right. Something went wrong in log 347. If you can send us the code and the logs we can take a look. See you, Klaus. On 6/18/07, Travelmole Travelmole wrote: > Hi, > > I have taken the snapshot, but 347 snapshot has been created. > > -rw-r--r-- 1 root root 197725 Jun 18 10:27 > 000000000000000000347.snapshot > > Will this snapshot be having all the datas? > > Pls sugess me the next step to do. > > Thanks, > Om > > > Travelmole Travelmole wrote: > Hi Klaus, > > Thanks for responce. > > Could you pls sugess me the step by step way to take a complete snapshot and > restore. > > Thanks & Regards, > Om > Klaus Wuestefeld wrote: > Hi :) > > Simply take a snapshot and delete the log. That should do it. > > But can you send us the log and the code? I have never seen this case. > > Thanks, Klaus. > > > > On 6/16/07, Travelmole Travelmole wrote: > > Hi everybody, > > > > I am using prevayler version 2.1 to store my data. The application has > been > > running for last 2 years. > > > > Since last month, new data created is being lost when we restart the web > > server. > > > > In the catalina.out file, I am seeing multiple instances of this exception > - > > > > java.io.IOException: Attempt to create command log file that already > > existed: /home/demo/prevayler/pro > > jects/000000000000000000348.commandLog > > at > > > org.prevayler.implementation.NumberFileCreator.newLog(NumberFileCreator.java:28) > > at > > > org.prevayler.implementation.CommandOutputStream.logStream(CommandOutputStream.java:52) > > at > > > org.prevayler.implementation.CommandOutputStream.logStream(CommandOutputStream.java:58) > > at > > > org.prevayler.implementation.CommandOutputStream.writeCommand(CommandOutputStream.java:26) > > at > > > org.prevayler.implementation.SnapshotPrevayler.executeCommand(SnapshotPrevayler.java:71) > > .... > > .... > > > > I am using linux, Tomcat as web server. > > > > List of last few CommandLog file is as below ( note the sequence of file > is > > not in same sequence as dates !! ) - > > > > -rwxrw-rw- 1 root root 100554 Feb 8 03:08 > > 000000000000000000341.commandLog > > -rwxrw-rw- 1 root root 100490 Feb 8 03:08 > > 000000000000000000342.commandLog > > -rwxrw-rw- 1 root root 59599 Feb 8 03:08 > > 000000000000000000343.commandLog > > -rwxrw-rw- 1 root root 101121 May 25 08:35 > > 000000000000000000344.commandLog > > -rwxrw-rw- 1 root root 100992 May 25 08:35 > > 000000000000000000345.commandLog > > -rwxrw-rw- 1 root root 100690 May 25 08:35 > > 000000000000000000346.commandLog > > -rwxrw-rw- 1 root root 100901 Jun 8 05:26 > > 000000000000000000347.commandLog > > -rwxrw-rw- 1 root root 101121 May 25 09:06 > > 000000000000000000348.commandLog > > -rwxrw-rw- 1 root root 100992 May 25 09:06 > > 000000000000000000349.commandLog > > -rwxrw-rw- 1 root root 100690 May 25 09:06 > > 000000000000000000350.commandLog > > -rwxrw-rw- 1 root root 58317 May 25 09:06 > > 000000000000000000351.commandLog > > > > How do I resolve this problem ? > > > > Thanks & Regards, > > Om > > > > ________________________________ > > Be a PS3 game guru. > > Get your game face on with the latest PS3 news and previews at Yahoo! > Games. > > > > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by DB2 Express > > Download DB2 Express C - the FREE version of DB2 express and take > > control of your XML. No limits. Just data. Click to get it now. > > http://sourceforge.net/powerbar/db2/ > > _______________________________________________ > > Prevayler-coders mailing list > > Pre...@li... > > > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > ________________________________ > Looking for a deal? Find great prices on flights and hotels with Yahoo! > FareChase.------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/_______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > ________________________________ > Yahoo! oneSearch: Finally, mobile search that gives answers, not web links. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Prevayler-coders mailing list Pre...@li... https://lists.sourceforge.net/lists/listinfo/prevayler-coders --------------------------------- Be a PS3 game guru. Get your game face on with the latest PS3 news and previews at Yahoo! Games. |
|
From: Klaus W. <kla...@gm...> - 2007-06-19 03:56:40
|
No. All transactions after log 347 will not be in this snapshot. Brians guess is probably right. Something went wrong in log 347. If you can send us the code and the logs we can take a look. See you, Klaus. On 6/18/07, Travelmole Travelmole <tes...@ya...> wrote: > Hi, > > I have taken the snapshot, but 347 snapshot has been created. > > -rw-r--r-- 1 root root 197725 Jun 18 10:27 > 000000000000000000347.snapshot > > Will this snapshot be having all the datas? > > Pls sugess me the next step to do. > > Thanks, > Om > > > Travelmole Travelmole <tes...@ya...> wrote: > Hi Klaus, > > Thanks for responce. > > Could you pls sugess me the step by step way to take a complete snapshot and > restore. > > Thanks & Regards, > Om > Klaus Wuestefeld <kla...@gm...> wrote: > Hi :) > > Simply take a snapshot and delete the log. That should do it. > > But can you send us the log and the code? I have never seen this case. > > Thanks, Klaus. > > > > On 6/16/07, Travelmole Travelmole wrote: > > Hi everybody, > > > > I am using prevayler version 2.1 to store my data. The application has > been > > running for last 2 years. > > > > Since last month, new data created is being lost when we restart the web > > server. > > > > In the catalina.out file, I am seeing multiple instances of this exception > - > > > > java.io.IOException: Attempt to create command log file that already > > existed: /home/demo/prevayler/pro > > jects/000000000000000000348.commandLog > > at > > > org.prevayler.implementation.NumberFileCreator.newLog(NumberFileCreator.java:28) > > at > > > org.prevayler.implementation.CommandOutputStream.logStream(CommandOutputStream.java:52) > > at > > > org.prevayler.implementation.CommandOutputStream.logStream(CommandOutputStream.java:58) > > at > > > org.prevayler.implementation.CommandOutputStream.writeCommand(CommandOutputStream.java:26) > > at > > > org.prevayler.implementation.SnapshotPrevayler.executeCommand(SnapshotPrevayler.java:71) > > .... > > .... > > > > I am using linux, Tomcat as web server. > > > > List of last few CommandLog file is as below ( note the sequence of file > is > > not in same sequence as dates !! ) - > > > > -rwxrw-rw- 1 root root 100554 Feb 8 03:08 > > 000000000000000000341.commandLog > > -rwxrw-rw- 1 root root 100490 Feb 8 03:08 > > 000000000000000000342.commandLog > > -rwxrw-rw- 1 root root 59599 Feb 8 03:08 > > 000000000000000000343.commandLog > > -rwxrw-rw- 1 root root 101121 May 25 08:35 > > 000000000000000000344.commandLog > > -rwxrw-rw- 1 root root 100992 May 25 08:35 > > 000000000000000000345.commandLog > > -rwxrw-rw- 1 root root 100690 May 25 08:35 > > 000000000000000000346.commandLog > > -rwxrw-rw- 1 root root 100901 Jun 8 05:26 > > 000000000000000000347.commandLog > > -rwxrw-rw- 1 root root 101121 May 25 09:06 > > 000000000000000000348.commandLog > > -rwxrw-rw- 1 root root 100992 May 25 09:06 > > 000000000000000000349.commandLog > > -rwxrw-rw- 1 root root 100690 May 25 09:06 > > 000000000000000000350.commandLog > > -rwxrw-rw- 1 root root 58317 May 25 09:06 > > 000000000000000000351.commandLog > > > > How do I resolve this problem ? > > > > Thanks & Regards, > > Om > > > > ________________________________ > > Be a PS3 game guru. > > Get your game face on with the latest PS3 news and previews at Yahoo! > Games. > > > > > > > ------------------------------------------------------------------------- > > This SF.net email is sponsored by DB2 Express > > Download DB2 Express C - the FREE version of DB2 express and take > > control of your XML. No limits. Just data. Click to get it now. > > http://sourceforge.net/powerbar/db2/ > > _______________________________________________ > > Prevayler-coders mailing list > > Pre...@li... > > > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > ________________________________ > Looking for a deal? Find great prices on flights and hotels with Yahoo! > FareChase.------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/_______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > > > ________________________________ > Yahoo! oneSearch: Finally, mobile search that gives answers, not web links. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > |
|
From: Travelmole T. <tes...@ya...> - 2007-06-18 15:11:47
|
Hi, I have taken the snapshot, but 347 snapshot has been created. -rw-r--r-- 1 root root 197725 Jun 18 10:27 000000000000000000347.snapshot Will this snapshot be having all the datas? Pls sugess me the next step to do. Thanks, Om Travelmole Travelmole <tes...@ya...> wrote: Hi Klaus, Thanks for responce. Could you pls sugess me the step by step way to take a complete snapshot and restore. Thanks & Regards, Om Klaus Wuestefeld <kla...@gm...> wrote: Hi :) Simply take a snapshot and delete the log. That should do it. But can you send us the log and the code? I have never seen this case. Thanks, Klaus. On 6/16/07, Travelmole Travelmole wrote: > Hi everybody, > > I am using prevayler version 2.1 to store my data. The application has been > running for last 2 years. > > Since last month, new data created is being lost when we restart the web > server. > > In the catalina.out file, I am seeing multiple instances of this exception - > > java.io.IOException: Attempt to create command log file that already > existed: /home/demo/prevayler/pro > jects/000000000000000000348.commandLog > at > org.prevayler.implementation.NumberFileCreator.newLog(NumberFileCreator.java:28) > at > org.prevayler.implementation.CommandOutputStream.logStream(CommandOutputStream.java:52) > at > org.prevayler.implementation.CommandOutputStream.logStream(CommandOutputStream.java:58) > at > org.prevayler.implementation.CommandOutputStream.writeCommand(CommandOutputStream.java:26) > at > org.prevayler.implementation.SnapshotPrevayler.executeCommand(SnapshotPrevayler.java:71) > .... > .... > > I am using linux, Tomcat as web server. > > List of last few CommandLog file is as below ( note the sequence of file is > not in same sequence as dates !! ) - > > -rwxrw-rw- 1 root root 100554 Feb 8 03:08 > 000000000000000000341.commandLog > -rwxrw-rw- 1 root root 100490 Feb 8 03:08 > 000000000000000000342.commandLog > -rwxrw-rw- 1 root root 59599 Feb 8 03:08 > 000000000000000000343.commandLog > -rwxrw-rw- 1 root root 101121 May 25 08:35 > 000000000000000000344.commandLog > -rwxrw-rw- 1 root root 100992 May 25 08:35 > 000000000000000000345.commandLog > -rwxrw-rw- 1 root root 100690 May 25 08:35 > 000000000000000000346.commandLog > -rwxrw-rw- 1 root root 100901 Jun 8 05:26 > 000000000000000000347.commandLog > -rwxrw-rw- 1 root root 101121 May 25 09:06 > 000000000000000000348.commandLog > -rwxrw-rw- 1 root root 100992 May 25 09:06 > 000000000000000000349.commandLog > -rwxrw-rw- 1 root root 100690 May 25 09:06 > 000000000000000000350.commandLog > -rwxrw-rw- 1 root root 58317 May 25 09:06 > 000000000000000000351.commandLog > > How do I resolve this problem ? > > Thanks & Regards, > Om > > ________________________________ > Be a PS3 game guru. > Get your game face on with the latest PS3 news and previews at Yahoo! Games. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Prevayler-coders mailing list Pre...@li... https://lists.sourceforge.net/lists/listinfo/prevayler-coders --------------------------------- Looking for a deal? Find great prices on flights and hotels with Yahoo! FareChase.------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/_______________________________________________ Prevayler-coders mailing list Pre...@li... https://lists.sourceforge.net/lists/listinfo/prevayler-coders --------------------------------- Yahoo! oneSearch: Finally, mobile search that gives answers, not web links. |
|
From: Travelmole T. <tes...@ya...> - 2007-06-18 09:46:26
|
Hi Klaus, Thanks for responce. Could you pls sugess me the step by step way to take a complete snapshot and restore. Thanks & Regards, Om Klaus Wuestefeld <kla...@gm...> wrote: Hi :) Simply take a snapshot and delete the log. That should do it. But can you send us the log and the code? I have never seen this case. Thanks, Klaus. On 6/16/07, Travelmole Travelmole wrote: > Hi everybody, > > I am using prevayler version 2.1 to store my data. The application has been > running for last 2 years. > > Since last month, new data created is being lost when we restart the web > server. > > In the catalina.out file, I am seeing multiple instances of this exception - > > java.io.IOException: Attempt to create command log file that already > existed: /home/demo/prevayler/pro > jects/000000000000000000348.commandLog > at > org.prevayler.implementation.NumberFileCreator.newLog(NumberFileCreator.java:28) > at > org.prevayler.implementation.CommandOutputStream.logStream(CommandOutputStream.java:52) > at > org.prevayler.implementation.CommandOutputStream.logStream(CommandOutputStream.java:58) > at > org.prevayler.implementation.CommandOutputStream.writeCommand(CommandOutputStream.java:26) > at > org.prevayler.implementation.SnapshotPrevayler.executeCommand(SnapshotPrevayler.java:71) > .... > .... > > I am using linux, Tomcat as web server. > > List of last few CommandLog file is as below ( note the sequence of file is > not in same sequence as dates !! ) - > > -rwxrw-rw- 1 root root 100554 Feb 8 03:08 > 000000000000000000341.commandLog > -rwxrw-rw- 1 root root 100490 Feb 8 03:08 > 000000000000000000342.commandLog > -rwxrw-rw- 1 root root 59599 Feb 8 03:08 > 000000000000000000343.commandLog > -rwxrw-rw- 1 root root 101121 May 25 08:35 > 000000000000000000344.commandLog > -rwxrw-rw- 1 root root 100992 May 25 08:35 > 000000000000000000345.commandLog > -rwxrw-rw- 1 root root 100690 May 25 08:35 > 000000000000000000346.commandLog > -rwxrw-rw- 1 root root 100901 Jun 8 05:26 > 000000000000000000347.commandLog > -rwxrw-rw- 1 root root 101121 May 25 09:06 > 000000000000000000348.commandLog > -rwxrw-rw- 1 root root 100992 May 25 09:06 > 000000000000000000349.commandLog > -rwxrw-rw- 1 root root 100690 May 25 09:06 > 000000000000000000350.commandLog > -rwxrw-rw- 1 root root 58317 May 25 09:06 > 000000000000000000351.commandLog > > How do I resolve this problem ? > > Thanks & Regards, > Om > > ________________________________ > Be a PS3 game guru. > Get your game face on with the latest PS3 news and previews at Yahoo! Games. > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Prevayler-coders mailing list > Pre...@li... > https://lists.sourceforge.net/lists/listinfo/prevayler-coders > > ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Prevayler-coders mailing list Pre...@li... https://lists.sourceforge.net/lists/listinfo/prevayler-coders --------------------------------- Looking for a deal? Find great prices on flights and hotels with Yahoo! FareChase. |
