If MVCC is enabled, changing the lock mode (LOCK_MODE) has no effect. See also http://h2database.com/html/advanced.html#mvcc for details
在 2013年11月28日星期四UTC+8上午4时01分01秒，Gili写道： > > Hi Thomas, > > Thanks for the clarification. > > What happens when MVCC is on and a user sets the transaction isolation to > something other than read committed? > > Thanks, > Gili > > On 27/11/2013 2:22 PM, Thomas Mueller wrote: > > Hi, > > When MVCC is disabled, see > http://h2database.com/html/advanced.html#transaction_isolation > > When MVCC is enabled, then H2 behaves in the same way as with read > committed. At least currently. In the future this may change. See also > http://h2database.com/html/advanced.html#mvcc for details > > Regards, > Thomas > > Thomas, > > I just want to make sure that you don't lose track of this question > (since you replied to a follow-up discussion): what is the difference > between SERIALIZABLE isolation when MVCC is on or off? > > Thanks, > Gili > > On Sunday, November 24, 2013 10:50:18 AM UTC-5, Gili wrote: >> >> Hi Thomas, >> >> It was my understanding that SERIALIZABLE isolation in MVCC mode is >> supposed to behave like SERIALIZABLE_SNAPSHOT. At least, that's the case >> for Postgres. If SERIALIZABLE behaves the way you mention in MVCC mode, >> what is the difference if MVCC mode is on or off? It doesn't sound like >> there is a difference. >> >> Gili >> >> On 24/11/2013 7:49 AM, Thomas Mueller wrote: >> >> Hi, >> >> > How does SERIALIZABLE isolation in MVCC mode compare to >> SERIALIZABLE_SNAPSHOT? Aren't they the same thing? >> >> I can only refer to Google, which found this: >> http://blogs.msdn.com/b/craigfr/archive/2007/05/16/serializable-vs-snapshot-isolation-level.aspx<http://www.google.com/url?q=http%3A%2F%2Fblogs.msdn.com%2Fb%2Fcraigfr%2Farchive%2F2007%2F05%2F16%2Fserializable-vs-snapshot-isolation-level.aspx&sa=D&sntz=1&usg=AFQjCNHL8riZ1UQ_8q-_ir8sxbV5VX7_XQ> >> >> Regards, >> Thomas >> >> >> >> >> On Sun, Nov 24, 2013 at 1:10 PM, cowwoc <[email protected]<javascript:> >> > wrote: >> >>> Hi Thomas, >>> >>> Thanks for the reply. >>> >>> >>> On 21/11/2013 6:43 AM, Thomas Mueller wrote: >>> >>> > Does H2 implement Serializable Snapshot >>> >>> Currently not, and I didn't plan to support it. In theory, it could be >>> supported with the MVStore, but I don't currently think it's very important. >>> >>> >>> How does SERIALIZABLE isolation in MVCC mode compare to >>> SERIALIZABLE_SNAPSHOT? Aren't they the same thing? >>> >>> Thanks, >>> Gili >>> >>> >>> >>> Regards, >>> Thomas >>> >>> >>> >>> On Sat, Aug 24, 2013 at 8:10 PM, Gili <[email protected]<javascript:> >>> > wrote: >>> >>>> Thomas, >>>> >>>> Can you please clarify how H2's implementation of MVCC interacts with >>>> Transaction Isolation? >>>> >>>> 1. Is there a practical difference between READ_COMMITTED and >>>> REPEATABLE_READ in MVCC mode? Doesn't MVCC prevent READ_COMMITTED from >>>> seeing updates committed by other transactions (after the current >>>> transaction has already read the row once)? If so, aren't they >>>> identical? >>>> 2. Same question for REPEATABLE_READ and SERIALIZABLE isolation in >>>> MVCC mode. >>>> 3. Does H2 implement >>>> >>>> http://en.wikipedia.org/wiki/Serializable_Snapshot_Isolation#Serializable_Snapshot_Isolation<http://www.google.com/url?q=http%3A%2F%2Fen.wikipedia.org%2Fwiki%2FSerializable_Snapshot_Isolation%23Serializable_Snapshot_Isolation&sa=D&sntz=1&usg=AFQjCNExhEdVr75dgDxP5G40R7fHvGul7A> >>>> for >>>> MVCC? If not, do you plan on doing so in the future? >>>> >>>> Thank you, >>>> Gili >>>> >>>> >>>> On Sunday, June 23, 2013 11:23:16 AM UTC-4, Gili wrote: >>>>> >>>>> >>>>> Thanks Thomas. So you're saying I can use other isolation modes >>>>> with MVCC as I normally would? >>>>> >>>>> Gili >>>>> >>>>> On 23/06/2013 7:39 AM, Thomas Mueller wrote: >>>>> >>>>> Hi, >>>>> >>>>> H2 only supports row level locks when using the MVCC mode. By the >>>>> way, the plan is that in the future (with the MVStore), the MVCC mode >>>>> will >>>>> be the default mode. >>>>> >>>>> Insert/delete operations don't wait for other operations that do not >>>>> conflict. They wait for conflicting operations (changes on the same rows). >>>>> >>>>> > What happens in the following scenario? >>>>> >>>>> H2 uses "read committed" by default, so T1 will see the value from >>>>> T2. >>>>> >>>>> Regards, >>>>> Thomas >>>>> >>>>> >>>>> >>>>> >>>>> On Thu, Jun 20, 2013 at 8:18 PM, Gili >>>>> <[email protected]<javascript:> >>>>> > wrote: >>>>> >>>>>> Hi, >>>>>> >>>>>> Looking at >>>>>> http://www.h2database.com/html/advanced.html#mvcc<http://www.google.com/url?q=http%3A%2F%2Fwww.h2database.com%2Fhtml%2Fadvanced.html%23mvcc&sa=D&sntz=1&usg=AFQjCNEZaGOPQsTGn9GcPhWhhmbHy_80kw> >>>>>> , >>>>>> >>>>>> >>>>>> 1. Does H2 only support table locks or MVCC? Or is there a way to >>>>>> get row locks without MVCC? >>>>>> 2. Using MVCC, do insert/delete operations wait until all open >>>>>> transactions complete? >>>>>> 3. Does this imply that open transactions will never experience >>>>>> insert/delete rows by other threads in mid-transaction? >>>>>> 4. What happens in the following scenario? >>>>>> >>>>>> Database contains a single row: count[value=1] >>>>>> T1: Open transaction >>>>>> T2: Open transaction >>>>>> T2: update count set value=2 where value=1 >>>>>> T2: commit >>>>>> T1: select value from count >>>>>> >>>>>> Will T1 see a value of 1 or 2? >>>>>> >>>>>> Thanks, >>>>>> Gili >>>>>> -- >>>>>> You received this message because you are subscribed to the Google >>>>>> Groups "H2 Database" group. >>>>>> To unsubscribe from this group and stop receiving emails from it, >>>>>> send an email to [email protected] <javascript:>. >>>>>> To post to this group, send email to >>>>>> [email protected]<javascript:> >>>>>> . >>>>>> Visit this group at http://groups.google.com/group/h2-database. >>>>>> For more options, visit https://groups.google.com/groups/opt_out. >>>>>> >>>>>> >>>>>> >>>>> >>>>> -- >>>>> You received this message because you are subscribed to a topic in the >>>>> Google Groups "H2 Database" group. >>>>> To unsubscribe from this topic, visit >>>>> https://groups.google.com/d/topic/h2-database/bGNuMpLr8IY/unsubscribe. >>>>> To unsubscribe from this group and all its topics, send an email to >>>>> [email protected] <javascript:>. >>>>> To post to this group, send email to >>>>> [email protected]<javascript:> >>>>> . >>>>> Visit this group at http://groups.google.com/group/h2-database. >>>>> For more options, visit https://groups.google.com/groups/opt_out. >>>>> >>>>> >>>>> >>>>> >>>>> -- >>>> You received this message because you are subscribed to the Google >>>> Groups "H2 Database" group. >>>> To unsubscribe from this group and stop receiving emails from it, send >>>> an email to [email protected] <javascript:>. >>>> To post to this group, send email to [email protected]<javascript:> >>>> . >>>> Visit this group at http://groups.google.com/group/h2-database. >>>> For more options, visit https://groups.google.com/groups/opt_out. >>>> >>> >>> -- >>> You received this message because you are subscribed to a topic in the >>> Google Groups "H2 Database" group. >>> To unsubscribe from this topic, visit >>> https://groups.google.com/d/topic/h2-database/bGNuMpLr8IY/unsubscribe. >>> To unsubscribe from this group and all its topics, send an email to >>> [email protected] <javascript:>. >>> To post to this group, send email to [email protected]<javascript:> >>> . >>> Visit this group at http://groups.google.com/group/h2-database. >>> For more options, visit https://groups.google.com/groups/opt_out. >>> >>> >>> -- >>> You received this message because you are subscribed to the Google >>> Groups "H2 Database" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected] <javascript:>. >>> To post to this group, send email to [email protected]<javascript:> >>> . >>> Visit this group at http://groups.google.com/group/h2-database. >>> For more options, visit https://groups.google.com/groups/opt_out. >>> >> >> -- >> You received this message because you are subscribed to a topic in the >> Google Groups "H2 Database" group. >> To unsubscribe from this topic, visit >> https://groups.google.com/d/topic/h2-database/bGNuMpLr8IY/unsubscribe. >> To unsubscribe from this group and all its topics, send an email to >> [email protected] <javascript:>. >> To post to this group, send email to [email protected]<javascript:> >> . >> Visit this group at http://groups.google.com/group/h2-database. >> For more options, visit https://groups.google.com/groups/opt_out. >> >> >> -- > You received this message because you are subscribed to the Google Groups > "H2 Database" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected] <javascript:>. > To post to this group, send email to [email protected]<javascript:> > . > Visit this group at http://groups.google.com/group/h2-database. > For more options, visit https://groups.google.com/groups/opt_out. > -- > You received this message because you are subscribed to a topic in the > Google Groups "H2 Database" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/h2-database/bGNuMpLr8IY/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > [email protected] <javascript:>. > To post to this group, send email to [email protected]<javascript:> > . > Visit this group at http://groups.google.com/group/h2-database. > For more options, visit https://groups.google.com/groups/opt_out. > > > -- You received this message because you are subscribed to the Google Groups "H2 Database" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/h2-database. For more options, visit https://groups.google.com/groups/opt_out.
