Index of files created by PXC¶

GRA_\*.log These files contain binlog events in ROW format representing the failed transaction. That means that the replica thread was not able to apply one of the transactions. For each of those file, a corresponding warning or error message is present in the mysql error log file. Those error can also be false positives like a bad DDL statement (dropping a table that doesn’t exists for example) and therefore nothing to worry about. However it’s always recommended to check these log to understand what’s is happening. To be able to analyze these files binlog header needs to be added to the log file. To create the GRA_HEADER file you need an instance running with binlog_checksum set to NONE and extract first 120 bytes from the binlog file: This information can be used for checking the MySQL error log for the corresponding error message. ??? example “Error message”
```
  ```text
  160805  9:33:37 8:52:21 [ERROR] Slave SQL: Error 'Unknown table 'test'' on query. Default database: 'test'. Query: 'drop table test', Error_code: 1051
  160805  9:33:37 8:52:21 [Warning] WSREP: RBR event 1 Query apply warning: 1, 3
  ```
```
In this example DROP TABLE statement was executed on a table that doesn’t exist.

gcache.page See [gcache.page_size](wsrep-provider-index.md#gcachepage_size) !!! admonition “See also”

  [Percona Database Performance Blog: All You Need to Know About GCache (Galera-Cache) :octicons-link-external-16:](https://www.percona.com/blog/2016/11/16/all-you-need-to-know-about-gcache-galera-cache/)

galera.cache This file is used as a main writeset store. It’s implemented as a permanent ring-buffer file that is preallocated on disk when the node is initialized. File size can be controlled with the variable [gcache.size](wsrep-provider-index.md#gcachesize). If this value is bigger, more writesets are cached and chances are better that the re-joining node will get IST instead of SST. Filename can be changed with the [gcache.name](wsrep-provider-index.md#gcachename) variable.

`## grastate.dat`¶

This file contains the Galera state information.

version - grastate version
uuid - a unique identifier for the state and the sequence of changes it undergoes.For more information on how UUID is generated see UUID.
seqno - Ordinal Sequence Number, a 64-bit signed integer used to denote the position of the change in the sequence. seqno is 0 when no writesets have been generated or applied on that node, i.e., not applied/generated across the lifetime of a grastate file. -1 is a special value for the seqno that is kept in the grastate.dat while the server is running to allow Galera to distinguish between a clean and an unclean shutdown. Upon a clean shutdown, the correct seqno value is written to the file. So, when the server is brought back up, if the value is still -1 , this means that the server did not shut down cleanly. If the value is greater than 0, this means that the shutdown was clean. -1 is then written again to the file in order to allow the server to correctly detect if the next shutdown was clean in the same manner.
cert_index - cert index restore through grastate is not implemented yet

Examples of this file look like this:

In case server node has this state when not running it means that that node crashed during the transaction processing.

GALERA saved state
version: 2.1
uuid:    1917033b-7081-11e2-0800-707f5d3b106b
seqno:   -1
cert_index:

In case server node has this state when not running it means that the node was gracefully shut down.

GALERA saved state
version: 2.1
uuid:    1917033b-7081-11e2-0800-707f5d3b106b
seqno:   5192193423942
cert_index:

In case server node has this state when not running it means that the node crashed during the DDL. The same zero UUID and seqno of -1 also appear after the cluster votes the node out for a data inconsistency: Galera marks the state corrupt so you can investigate while the evicted node is still running.

GALERA saved state
version: 2.1
uuid:    00000000-0000-0000-0000-000000000000
seqno:   -1
cert_index:

A corrupt grastate.dat does not force a full SST on the next start. Restart recovery reads a usable position from InnoDB and the node can rejoin with IST, which leaves the inconsistent dataset in place. To force SST after an inconsistency eviction, enable [repl.force_sst_after_inconsistency](wsrep-provider-index.md#replforce_sst_after_inconsistency) (the default is no, which leaves the file in place for inspection) or remove grastate.dat manually before restart. When that option is yes, the node deletes the file when it marks the state corrupt; copy the file first if you still need it.

gvwstate.dat This file is used for Primary Component recovery feature. This file is created once primary component is formed or changed, so you can get the latest primary component this node was in. And this file is deleted when the node is shutdown gracefully. First part contains the node UUID information. Second part contains the view information. View information is written between #vwbeg and #vwend. View information consists of:
view_id: [view_type] [view_uuid] [view_seq]. - view_type is always 3 which means primary view. view_uuid and view_seq identifies a unique view, which could be perceived as identifier of this primary component.
bootstrap: [bootstarp_or_not]. - it could be 0 or 1, but it does not affect primary component recovery process now.

member: [node’s uuid] [node’s segment]. - it represents all nodes in this primary component. ??? example “Example of the file”

  ```text
  my_uuid: c5d5d990-30ee-11e4-aab1-46d0ed84b408
  #vwbeg
  view_id: 3 bc85bd53-31ac-11e4-9895-1f2ce13f2542 2 
  bootstrap: 0
  member: bc85bd53-31ac-11e4-9895-1f2ce13f2542 0
  member: c5d5d990-30ee-11e4-aab1-46d0ed84b408 0
  #vwend
  ```

Index of files created by PXC¶

## grastate.dat¶

`## grastate.dat`¶